enabling technologies for organizational learningweb.mit.edu/smadnick/www/wp2/1993-05.pdf ·...

63
Enabling Technologies for Organizational Learning May 1993 CISL WP# 93-05 Steven Wan Pok Wu Stuart Madnick Stuart Madnick, E53-321 Sloan School of Management Massachusetts Institute of Technology Cambridge, MA 02138

Upload: phamnga

Post on 25-Jun-2018

213 views

Category:

Documents


0 download

TRANSCRIPT

Enabling Technologies forOrganizational Learning

May 1993 CISL WP# 93-05

Steven Wan Pok WuStuart Madnick

Stuart Madnick, E53-321

Sloan School of ManagementMassachusetts Institute of Technology

Cambridge, MA 02138

Table of Contents

PageCABSTRACT................................................................................. 1

IPart.. ................................................................................. 2

Chapter 1: An Introduction to Networked Organizations ............................ 31.1 Networking in the Real W orld ............................................................................ 31.2 Definitions and Attributes ................................................................................... 41.3 Drivers for Organizational Change and Improvement .................................... 51.4 Supportive Cultures.............................................................................................. 61.5 Enabling Technologies............................................................................................ 61.6 Evolution of Networked Organizations.............................................................. 71.7 D iscussion ................................................................................................................ 8

Chapter 2: Organizational Learning in Networked Organizations ....... 82.1 Learning Organizations .................................................................. 82.2 Individual and Organizational Learning.............................................. 92.3 Learning Goals............................................................................. 112.4 Learning Processes and Disciplines....................................................1

2.4.1 Learning Loops.................................................................. 112.4.2 Learning Disciplines............................................................ 13

2.5 Organizational Learning in practice ...................................................... 152.5.1 Planning as Learning ........................................................... 152.5.2 Learning in R&D Processes.................................................... 152.5.3 Total Quality Management.................................................... 162.5.4 Modeling as Learning........................................................... 162.5.5 Benchmarking ................................................................... 172.5.6 Experiential Learning........................................................... 17

2.6 Learning Disabilities and Impediments ............................................... 182.7 Impacts on Networked Organizations................................................. 192.8 Discussion .................................................................................. 20

Chapter 3: Organizational Learning in IT Processes.............................. 213.1 An Overview of IT Issues and Challenges..................... 213.2 Shared Responsibility for IT............................................................. 223.3 Information Politics ....................................................................... 233.4 The Organization as a Learning Laboratory.......................................... 233.5 Knowledge Creation in IT Processes................................................... 243.6 A Cyclic Learning Model ................................................................ 253.7 Discussion .................................................................................. 27

Summary of Part I............................................................................. 27

PartiHl ....................................................................................... 29

Chapter 4: Using E-mail ................................................................... 304.1 Overview of Electronic Mail............................................................. 304.2 Examples of E-Mail Interactions........................................................ 314.3 Sociology of E-mail ....................................................................... 324.4 Impact on Organizational Learning.................................................... 33

4.4.1 E-mail and Cyclic Learning.................................................... 344.4.2 E-mail and Learning Loops .................................................... 344.4.3 E-mail and the Learning Disciplines ......................................... 35

4.5 Impediments to Learning ................................................................ 364.6 Discussion .................................................................................. 36

Chapter 5: Implementing Groupware...........................375.1 Overview of Groupware and Collaborative Computing........................... 375.2 Examples of Groupware applications ................................................. 395.3 Organizational and cultural issues ..................................................... 405.4 Impact on Organizational Learning.................................................... 41

5.4.1 Groupware and Cyclic Learning ............................................. 415.4.2 Groupware and the Learning Loops ......................................... 425.4.3 Groupware and the Learning Disciplines................................... 43

5.5 Impediments to Learning ................................................................ 435.6 Discussion .................................................................................. 44

Chapter 6: IT STRATEGIC PLANNING ............................................ 446.1 Information Technology and Paradigm Shifts........................................ 446.2 Overview of IT Strategic Planning...................................................... 456.3 Two Models of IT infrastructure........................................................ 476.4 Impact on Organizational Learning.................................................... 48

6.4.1 IT Planning and Cyclic Learning ............................................. 486.4.2 IT Planning and the Learning Loops ........................................ 49

6.5 Impediments to Learning ................................................................ 526.6 Discussion .................................................................................. 52

Summary of Part II............................................................................ 53I. Review..................................................................................... 55II. Becoming a Learning Organization ............................................................ 56

Bibliography ands................. ............................................................ 59.......4

ENABLING TECHNOLOGIES FORORGANIZATIONAL LEARNING

by

STEVEN WAN POK WU

ABSTRACT

Networked organizations are becoming a trend particularly for globalization strategies. Theorganization's resources and capabilities are extended by a network of suppliers, customers andpartners. Internal and external relationships have to be connected with the organization bymeans of effective communication media and synergistic work processes. To sustain theircompetitive advantage, networked organizations have to learn continuously. They have tobecome learning organizations.

Organizational learning is more than just the collective sum of individual learning within anorganization. Learning enhances the capacity of the organization to adapt and innovate across allof the units along the entire value chain. Much of organizational learning takes place throughexperience. This experiential learning is formalized in the Problem-oriented learning cycle whichis a four-stage cyclic process : doing, reflecting, thinking, deciding and (re)doing. Learningprocesses can be classified as single loop or adaptive learning for improvement, double loop orgenerative learning for renewal, and triple loop learning for development. Finally, learning isunderpinned by the five disciplines defined by Peter Senge : personal mastery, mental models,team learning, shared vision, and systems thinking.

Information technology, an essential fixture in any modern organizational infrastructure,provides an almost ubiquitous setting for learning. Electronic mail is becoming an almostomnipresent communication medium linking workers around the world with round-the-clockmessaging capability. Groupware, particularly workflow automation, impacts organizationalbehavior and culture through its implementation. Information infrastructure planning provides acommon platform for cross-disciplinary cooperation in building a shared vision.

The goal of this thesis is to examine how using, implementing, and planning for, informationtechnology provide opportunities for learning, what are involved in the learning process andhow impediments to learning may be overcome.

Thesis Supervisor : Professor Stuart E. MadnickTitle : John Norris Maguire Professor of Information Technology

Part I

Part I explains the need for organizational learning, especially in corporations of the future.These corporations are envisioned to have horizontal or networked structures. Learningorganizations are those that are capable of creating knowledge and thus able to enhance theircompetitiveness by adding unique values to their products and services. Information andknowledge are perceived to be indispensable assets in the networked, knowledge-basedorganization. Information technology (IT) provides the means to communicate, coordinate andcooperate between the various components - employees, internal divisions, suppliers, customersand partners - of the extended networked organization. Hence, understanding the extent that ITprocesses support organizational learning in networked corporations and how learning enhancesIT processes is a crucial step towards realizing an effective learning environment.

Chapter One discusses the problems encountered by hierarchical organizations and themotivations to develop network structures. The likely corporate culture is team-oriented asopposed to individual-oriented. The critical success factor of the organization of the future is theability to learn. Training, appraisal and reward systems should reflect this orientation.Information technology is the key enabling technology which encompasses extensive computernetworks that support intra- as well as inter-organizational applications.

Chapter Two provides a broad overview of organizational learning. The learning organization isintroduced and learning goals are proposed. Individual learning is contrasted withorganizational learning. Adaptive (single loop) versus generative (double loop) learning arediscussed. Peter Senge's five learning disciplines are presented. Organizational learning inpractice may be embodied in a wide range of corporate group activities, from strategic planningto total quality management. Learning disabilities and ways to overcome them are examined.The impact of learning on the networked corporation is highlighted.

Chapter Three presents the case for IT as a vehicle for organizational learning. IT-enabledlearning covers a broad spectrum of processes. Using, implementing and planning are threemajor processes discussed here that impact various levels of employees in different action-thinking modes. To share their operational knowledge, line managers should form an ITmanagement architecture together with IT managers. They should collaborate through the rightinformation politics - federalism. For IT to be an effective vehicle for learning, the networkedorganization should be viewed as a learning system in which interactions between the ITmanagement architecture, technologies, projects and external relations result in knowledgecreation. The Problem-oriented Cyclic Learning Model is presented here as the core model fordiscussing how IT processes support, and are impacted by, learning.

Chapter 1: An Introduction to Networked Organizations

1.1 Networking in the Real World

The start of this decade has seen the fall of several titans of American business. The ailments thatbrought down General Motors, IBM, Westinghouse and Sears are principally due to the ineptnessof management to adapt to rapidly changing economic, business and technological landscapes.While these corporations may yet rise again from their present woes, a great deal has beendiscussed and written about their attempts at recovery.

Of particular interest is the path that IBM, an epitome of the hierarchical, monolithicorganization, has taken in re-structuring itself into thirteen autonomous operating companies. Inso doing, it is acknowledging the need for a radical transformation in order to be betterpositioned for the challenges of the nineties. Effectively, IBM has taken the first steps in evolvinginto a more flexible, anti-bureaucratic corporation in order to nurture a entrepreneurial climatefor each of its operating companies. In other words, it is tilting towards becoming a networkedcorporation. A major paradigm shift in organizational transformation has occurred.

According to James R. Houghton, Chairman of Corning, Inc., a diversified corporation that hasembodied some of the most common notions of a networked organization:

A network is an interrelated group of businesses with a wide range of ownershipstructures...Within each sector there are a variety of business structures that range fromtraditional line divisions to wholly-owned subsidiaries and alliances with othercompanies...A network is egalitarian. There is no parent company. A corporate staff isno more, or less, important than a line organization group. And being part of a jointventure is just as important as working at the hub of the network. [CORNINC89]

The above definition suggests that a truly networked corporation integrates not only theorganization's employees and resources, but also partners, suppliers, distributors, retailers andeven customers, who are dispersed across cultures, space and time. The obvious fruits of suchnetworking are global sharing of resources and expertise, worldwide customer service,production economies and leverage with suppliers.

For a more European flavor, consider ABB Asea Brown Boveri, a global organization of widebusiness diversity that resulted from a merger of two major corporations, ASEA and BrownBoveri. Led by Percy Barnevik, an advocate of strict decentralization, the organizing principlesare straightforward.

Along one dimension, the company is a distributed global network. Executives around the worldmake decisions on product strategies and performance without regard for national borders.Along a second dimension, it is a collection of traditionally organized national companies, eachserving its home market as effectively as possible. ABB's global matrix holds the two dimensionstogether. Barnevik refers to this as a "multi-domestic" enterprise. ABB is thus "global and local,big and small, radically decentralized with centralized reporting." [ABB-HBS91]

Japanese keiretsus are typically organized in a ring topology, a roundtable of equals. Mitsubishi,for example, is a collection of hundreds of interdependent companies strung together by cross-shareholdings, interlocking directorates, joint ventures, and long-term business relationships. Nosingle company predominates. Group members compete fiercely in the markets and with eachother. However, financial cross-holdings create a dense fabric of relationships that can beexploited when mutually beneficial. Such a safety net encourages long-term investments and

high-tech risk-taking. Networking is also exploited to collect, collate and integrate informationfrom all over the world. This sharing of a critical resource gives the keiretsu a significantheadstart on the competition.

There is no single model of the networked corporation. One has to consider historical, national,political, economic and business factors to understand the evolution of these organizations totheir present forms.

Downsizing or "delayering" has resulted in much of the middle management in US firms beingtrimmed, for reasons such as cost-cutting, improving responsiveness, the realization ofdiseconomies of scale. The worldwide merger-and-acquisition fever of the eighties too have hada deep impact on organizational re-structuring. Although a networked structure may notnecessarily have been intended, the stage has nonetheless been set for its subsequent refinement.While there is no flagbearer for networked corporations, some patterns may be discerned in theirmanagement practices and corporate cultures.

1.2 Definitions and Attributes

A more formal analysis of the key attributes of a networked approach to inter- and intra-organizational activities has been made by John F. Rockart and James E. Short [ROCKART91].Networks are defined as "one part of the firm's overall system of interrelationships to accomplishwork".

The study listed seven key attributes of the networked approach:

1) Shared goals. Networks typically organize around shared goals or objectives.2) Shared expertise. Networks allow for the sharing of expertise and knowledge across firms.3) Shared work. Networks allow for the sharing of work across groups not normally part of thestructure.4) Shared decision making. Networks allow for shared decision making, mainly throughenhanced access to critical information across the firm.5) Shared timing and issue prioritization. Networks allow for, and depend on, sharedprioritization and time horizons for critical issues and action steps.6) Shared responsibility. accountability, and trust. Networks depend on the sharing ofresponsibility, accountability, and trust in the organization.7) Shared recognition and reward. Shared rewards and shared recognition for cooperative workare implicit in the effective functioning of networks.

A note of caution is in place here. While these attributes are comprehensive, they are also verybroadly defined in rather idealized terms. The politics of sharing work, expertise, etc inevitablygive rise to internal conflicts. Conflict is also inherent in trying to integrate the wide range ofspecialized functions and skills. Anecdotal accounts also point to a tendency toward mutualdistrust amongst the members, in addition to self-interest on the part of the member companyover the overall corporate interests. This parochialism partly arises from the fact that internalmeasurements, appraisals, and incentives are mostly locally oriented. Thus, in evaluating theeffectiveness of networks, it is necessary to assess the quality of relationships, not just structuresand processes. A well-managed networked organization must create the necessary infrastructureto enable cross-boundary communication and information sharing.

In general, networks have either a flat hierarchy or none at all. Flexible but yet complex, they areadaptive - capable of being quickly reconfigured to meet changing needs. Leadership innetworks tends to be based on competence and personality rather than social or organizationalrank. The Chief Executive Officer is likely to assume the role of premier generalist-visionary-communicator. Such a role has been likened to that of an orchestra conductor. Power turns over

more frequently and more easily than in a bureaucracy, changing hands as new situations arisethat demand new skills [TOFFLER90, DAVIDOW92].

1.3 Drivers for Organizational Change and Improvement

The business environment of the coming decade is likely to be characterized by rapid growth inglobalization, volatile political environments, parochialism of regional trading blocs, additionallegislative and environmental requirements, governmental intervention in commerce, as well as anarrowing of the competence gap between current leaders and laggards particularly in processtechnology. The individual firm will be hard put to excel across all the dimensions necessary formaintaining a significant competitive advantage. The most likely response would be to increaselinkages with other firms, eventually resulting in a network.

Rockart and Short [ROCKART91] also discusses the organizational and business forces that drivethe move to networked organizations. These forces include:

1) Time to Market. "Time to market" refers to the firm's ability to design, produce, and bring tomarket new products quickly, and to better manage existing product lines. Compressing time tomarket requires increased integration of effort across functional departments.

2) Service. The need to improve service by drawing upon organizational-wide knowledge of thecustomer's status, current problems and anticipated problems drives the need for integrationacross many levels of the organization. Several divisions can work together to provide a total,seamless solution to the customer.

3) Ouality. Quality spans both product and process quality - the accuracy of the end productdesign and manufacturing processes as well as the degree of conformance to the customer'sneeds. Total quality management requires a concerted, inter-functional coordination andcooperation to achieve.

4) Risk. Market volatility, business complexity and competitive pressures all contribute toincreased risks in a number of ways. Strategic alliances help to spread the risks and share thefinancial burden of new technology R&D and product development.

5) Cost. Cost reductions are always a paramount concern in all organizations. Recent strategiesinclude outsourcing and consolidation of shared activities across subunits into one or a fewcentralized operations.

6) Partnership. The firm's external links with suppliers, customers and other firms will beenabled by communication technologies such as electronic data exchange (EDI), value-addednetworks, partnership programs and strategic alliances.

A business enterprise will stand to lose its competitiveness by ignoring the dynamics of theseforces. Increasingly, a business organization's financial performance is tied to that of its tradingpartners. There is therefore a perceived need to define, track and manage joint performanceacross these partnerships and alliances.

A networked organization provides a flexible structure for entering into alliances with severaltypes of organizations concurrently. The flattened structure allows for adaptability andresponsiveness to the market. The simpler management hierarchy encourages employeeparticipation in decision-making. Furthermore, the availability of comprehensive informationtechnology for supporting such inter-firm activities adds impetus to the formation of networks.

1.4 Supportive Cultures

Corporate culture, when well conceived and applied, will play a crucial role in conveying ashared purpose and vision to the employees. A corporate culture involves that role played by anorganization's values, symbols, rites and rituals in determining its overall performance. Thecorporate culture is shaped by its business environment. Hence, it unlikely to be successfullycreated by the brute force approach of putting together a set of idealized values and practices.Rather, to be effective, it has to be aligned with the organization's goals and to evolve with theorganization.

McKinsey and Co., the management consulting firm, has drawn up a blueprint for a horizontal(networked) company [FORTUNE,MAY92]. The key elements that directly bear on corporateculture include (1) making teams, not individuals, the focus of organizational performance anddesign; (2) combining managerial and non-managerial activities as often as possible, e.g. letworker teams take on hiring, evaluating and scheduling; (3) informing and training people on ajust-in-time, need-to-perform basis; (4) maximizing supplier and customer contact with everyonein the organization, e.g. by using joint problem-solving teams for all employees all the time; and(5) rewarding individual skill development and team performance instead of individualperformance alone. The essence of these practices is to emphasize team values over individualvalues, empower the non-managerial workers, continuously develop employees and encouragecross-boundary problem-solving.

However, the overriding critical success factor in building and exploiting network strengthswould be organizational learning - how the organization teaches its workers and creates anenvironment that enables learning. Says Harvard Professor Shoshanna Zuboff, author of The Ageof the Smart Machine: :

The twenty-first century company has to promote and nurture the capacity to improveand to innovate. That idea has radical implications. It means learning becomes the axialprinciple of organizations. It replaces control as the fundamental job of management."[FORTUNE,MAY92]

As the organizational form most likely to predominate in the coming decade, networkedorganizations will do well by fostering a climate in which learning at all levels is rewarded. Anetwork also has some built-in advantages in providing a ready-made platform for cross-organizational and cross-disciplinary learning. Industrial partners are enabled to shareexperiences and learn from each other on issues of common concern. Specialists from severalrelated disciplines can be brought together, electronically or otherwise, to focus on complexproblems.

1.5 Enabling Technologies

Without doubt, information networks will play a key role in enabling networked organizations.However, there is a wide spectrum of overlapping technologies presently available as well asintegrating architectures for building a cooperative, enterprise-wide electronic network.

For example, for interactive communication, technologies include electronic mail, shared-screenconferencing, video conferencing, and enhanced telephony. One-way communicationtechnologies include voice-mail and broadcasting. For task-oriented teams, groupwares thatenable collaborative work and improve productivity are multiplying in sophistication andversatility. Complex heterogeneous local-area as well as wide-area networks have also beenimplemented that tie together dozens of dis.parate information sources.

Fortunately, the advent of industry-wide consortia and user-driven enforcement of standards hasreduced an otherwise incomprehensible Tower-of-Babel of user interfaces, operating systems andnetwork protocols. While open systems are beginning to be adopted, much still needs to be donebefore a superhighway of connected but heterogeneous databases can be transparently accessedby users at any location within or even outside the organization.

The emphasis of networking goals should be connectivity on all dimensions - physical, logical,strategic and organizational. Physical connectivity refers to the ability to physically link andaccess information systems. Logical connectivity requires that data be merged and formatted intoa manageable form. Moreover, semantic conflicts - contradiction, ambiguity, incompleteness -have to be resolved by some embedded software sometimes referred to as knowbots. Strategicconnectivity refers the strategic benefits derived from easier, more efficient, and integratedorganizational access to information. Organizational connectivity enables synergy in terms ofincreased efficiency, improved productivity, shared learning and new business opportunities.

Zuboff coined the term "informated organization" to underscore the synergistic blending ofinformation technology with business processes. Such an informated organization will be alearning institution where learning is the heart of productive activity, or simply, a new form oflabor [ZUBOFF84].

1.6 Evolution of Networked Organizations

Organizational structure evolves as the firm grows from chaotic birth through rapid growth,slowing maturity and, possibly, precarious decline. During its early stage, the organization isinnovative and entrepreneurial, an exciting place and time to work. A key challenge is to retainand enhance the best of these characteristics as the organization diversifies and globalizes.Furthermore, the extended organization of today includes suppliers, customers and partners.New paradigms have to be developed to formalize and explain the new organizational structures.It is not uncommon to borrow metaphors from other disciplines, for example, biology andcomputer science.

In their book Paradigm Shift, Don Tapscott and Art Caston [TAPSCOTT93,p 331 adapt acomputing paradigm - client-server architecture - to define the open networked organization.Such an organization is based on cooperative, multi-disciplinary teams and businesses networkedtogether across the enterprise. Forsaking rigidity, it is a modular organization architecture inwhich business teams operate in a network of client and server functions. Teams are both clientand servers to other teams, which may be internal or external to the organization.

Looking ahead, Harvard economist Robert Reich foresees boundaries becoming so fluid thatcorporations become "temporary arrangements among entrepreneurial cadres". Except for high-volume, capital-intensive work, says Reich, "every big business will be a confederation of smallones. All small organizations will be constantly in the process of linking up into big ones."[FORTUNE,MAY92].

A more generalized network has been conceptualized in the form of the virtual corporation. Avirtual corporation is a temporary network of independent companies - suppliers, customers,even erstwhile rivals - linked by information technology to share skills, costs, and access to oneanother's markets. It will have neither central office nor organization chart. It will have nohierarchy, no vertical integration [BWEEK,FEB93]. No organization fitting the above notions ofvirtuality exists today. A more realistic example is the interfirm alliance forged by IBM, APPLEand Motorola to develop an operating system and a microprocessor for a new generation ofcomputers.

Whatever the eventual forms that the organization of the next decade evolves to, the chances arethat it will be horizontal, anti-bureaucratic, loose-knit, and orchestrated (as opposed to beingmanaged). It is one in which there is sharing of vision, resources, decision-making, work, andresponsibility. The organization of tomorrow, above all, will be oriented towards learning.

For the purpose of this thesis, the term networked corporation (or organization) will encompassall possible variants and forms (e.g., Peter Keen's relational organization; D. Quinn Mill's clusterorganization; Rosabeth Moss Kanter's "Dancing elephants"; Tom Peter's "life without hierarchy";Alvin Toffler's adhocracy; McKinsey's horizontal organization; or William H. Davidow's virtualorganization) that share these attributes. Through the protean interactions of its members, valueis added in terms of flexibility, innovation, entrepreneurship and responsiveness.

1.7 Discussion

In summary, networks provide a well-tested path for organizational evolution. However, theyshould not be created as ends in themselves but as means to some higher levels of organizationalimprovement and competitiveness. Above all, a network should not be contrived as a panaceafor corporate ills.

Information networks provide the backbone and nervous system for building networkedorganizations. In order for the networked corporation to realize its fullest potential and to be ableto derive the most benefits from its human resources, organizational learning will have to beaccorded a top management priority. With IT infrastructure becoming an integral component ofany modern corporation, it is all the more compelling to explore how planning, implementationand use of IT can directly support organizational learning.

Chapter 2: Organizational Learning in Networked Organizations

2.1 Learning Organizations

The "learning organization", is emerging as possibly the most influential managementparadigm of the nineties. By no means a new idea, organizational learning, once the turf ofserious researchers like Chris Argyris and Peter Senge, has become a subject of enthusiasticattention by popular writers like Tom Peters in his latest tome, Liberation Management.

As the decade unfolds, the command-and-control, multi-departmental, multi-divisional formconceived by Pierre S. du Pont and refined by Alfred P. Sloan is being supplanted by "theinformation-based organization, the organization of knowledge specialists." In drawingattention to this shift, management guru Peter Drucker [DRUCKER88] recognizes that thetoughest problem to be faced is that of ensuring the supply, preparation and testing of topmanagement people. There is a critical need for continuous learning that is proactive, shared,cross-disciplinary, and experiential. In addition, such learning has to prepare the manager forunforeseeable changes in the business environment by continuously challenging their currentmental models.

Drawing from his experience with Japanese corporations, Ikujiro Nonaka of HitotsubashiUniversity goes further in espousing the "knowledge-creating company". [NONAKA91] Thevision centers around creating a organization whose ability to continually improve itsprocesses and systems comes from continually enhancing its underlying knowledge base.Understanding the knowledge-creating company starts with rethinking the meaning ofknowledge and learning.

Nonaka argues that most of the organization's knowledge lies in tacit knowledge carried inthe heads of its members. Such knowledge is obtained by experimentation and action; itforms the "capacity for effective action". Organizational knowledge is not just hard, objectiveinformation but includes highly subjective insights, intuitions, beliefs and hunches ofemployees. Tacit knowledge is converted into explicit knowledge when an employeearticulates his vision. Explicit knowledge can be internalized into tacit knowledge when anemployee adapts his mental models as a result of learning. Nonaka further contends that:

A company is not a machine but a living organism. Much like an individual, it canhave a collective sense of identity and fundamental purpose. This is theorganizational equivalent of self knowledge - a shared understanding of what thecompany stands for, where it's going, what kind of world it wants to live in, and,most importantly, how it intends to make that world a reality...To create knowledgemeans quite literallyto re-create the company and everyone in it in a nonstop processof personal and organizational self-renewal.

According to Dr. Edward Deming, father of Total Quality Management, people are born withintrinsic motivation, self esteem, curiosity to learn, and joy in learning. However, piecemeal,ill-conceived approaches to managing workers such as management by objectives, employeeranking systems, incentive pays, bonus plans has led to the destruction of worker motivation.

The solution, according to Deming, lies in a deeper transformation that starts withdistinguishing "intrinsic from extrinsic motivation" and developing an organization whereeveryone, from top to bottom, is a learner.

Hence, the learning organization is also a knowledge-creating organization in which theindividual employee is able to find fulfillment through learning, creating and sharingknowledge. It differs from resource-based organizations in the fundamental tasks of directionsetting, nature of thinking, responsibility for thinking and executing, conflict resolution, androle of leadership. These tasks are further discussed under the section on "Learning Processesand Disciplines".

In order for effective learning to take place, all employees, including managers at all levels,must become active learners. Incentive, reward systems, norms, and'organizational structuresthat stimulate learning must be created. Organizational learning and change must be treatedas perpetual and inevitable processes. Learning must be imbedded into the corporate culture.

2.2 Individual and Organizational Learning

Learning can be defined as the process within the organization by which knowledge aboutaction-outcome relationships and the effects of the environment on these relationship aredeveloped [DUNCAN79]. The outcome of the learning process is knowledge that isdistributed across the organization, is communicable among members, has consensualvalidity, and is integrated into the working procedures of the organization [MEYER92, p 1691.

Edgar Schein says that the organization learns "when the capacity to adapt and innovate isshared across all of the units along the entire value chain, and the organization is able toimplement its capacity in actions that get the desired results." [SCHEIN92]

Peter Senge adds that "learning concerns the enhancement of the capacity to create." Reallearning occurs when people are trying to do something that they want to do. There are twoaspects to learning: doing and expanding one's capacity to do or create. [PETERMK92]

All organizations learn through explicitly defined processes or otherwise. For effectiveorganizational learning to take place, deliberate as opposed to unfocused effort is required inorder to achieved stated learning goals.

While current understanding individual learning is still inconclusive, that of organizationallearning is in its infancy. Organizations are composed of a collection of individuals and mustultimately learn via their individual members. A dilemma posed by Chris Argyris andDonald Schon [ARGYRIS78] concerns the observation that an organization is both dependenton individual learning and yet capable of its own learning:

There is something paradoxical here. Organizations are not merely collections ofindividuals, yet there are no organizations without such collections. Similarly,organizational learning is not merely individual learning, yet organizations learn onlythrough the experience and actions of individuals. What, then, are we to make oforganizational learning? What is an organization that it may learn?

However, Arie P. De Geus, former head of planning at Royal/Dutch Shell, notes thatorganizational learning is much more difficult than individual learning. The learning level ofthe team is often the lowest common denominator especially if the team members think ofthemselves as playing disconnected, specialized roles such as marketing, production and soon. The best learning therefore takes place in teams that accept that the whole is greater thanthe sum of the parts [GEUS88].

Arguing that to create shared understanding for implementation of new concepts of productdevelopment, learning has to be a team process, Steven C. Wheelwright and Kim B. Clark,authors of Revolutionizing Product Development, [WCLARK92,p293] state:

Team members bring different perspectives and different capabilities, and will readevidence in different ways. This can be a powerful and important source of insightand understanding. In effect, the kind of learning that is most crucial is learning thatcuts across the narrow, functionally oriented tasks in development and concentrateson the behavior of the development system. It is therefore important that theorganizational processes used to learn match the learning objectives...

Wheelwright and Clark also note that in order to capture insights from development projects,learning requires careful, systemic effort. In "learning by doing" engineers pay attention to thetask at hand, and develop ways to remember what they learn. Thus, without any explicitaction by management, individual learning takes place and know-how increases. This know-how is diffused within the work group through the use of stories, rules of thumb, changes inprocedures or methods or other forms of communication.

However, in a product development environment, tasks and capabilities cuts acrossfunctional boundaries. In additional to linkages between disciplines, functions anddepartments, there are complex decision and resource allocation processes. Individuals,however, tend to focus on the completion of specific tasks and activities over which they havecontrol.

At the development system level, where causal relationships may be intricately intertwined,the individual encounters difficulty in observing the interactions and even more so in drawingcomplex inferences about them. Furthermore, the natural incentives in an organization favorspressing ahead to the next project. The consequence is that most companies learn little fromtheir previous development experience.

Therefore, learning at the system level, and in particular about organizational processes thatdrive performance, requires systemic and tenacious effort. Besides gaining insights andimproved understanding, the organization must also determine how to change the process inorder to improve its performance.

Peter Senge [SENGE90,p1391 sums it up: "Organizations learn only through individuals wholearn. Individual learning does not guarantee organizational learning. But without it noorganizational learning occurs."

2.3 Learning Goals

We can roughly divide learning goals into two types - individual and organizational.

For the individual, the main goal is to improve the quality of his/her actions. The individualwants to develop a sense of personal mastery and feeling of connectedness to the group andorganization. Without being too conscious of it, the individual begins to overcome theimpediments that have blocked his learning ability as soon as he takes the first steps towardssharing knowledge with his or her peers.

Argyris offers a view of organizational learning as a process of detecting and correcting error,which he defines as "any feature of knowledge or knowing that inhibits learning." While mostindividuals and organizations are quite good at single loop learning, a more critical goal isdouble loop learning. These concepts are discussed under "learning loops" below.

According to Schein, organizational learning increases the capacity of the organization toadapt and to innovate in an increasingly turbulent and unpredictable world. [SCHEIN92].

Hiroyuki Itami, in his book, Mobilizing Invisible Assets, says that employees must be made tounderstand that the purpose of competitive strategy is to accumulate certain invisible assets -information-based resources like customer knowledge, management skills, etc. When a firmstarts to compete without the full complement of such assets, the employees are placed in astate of great creative tension. In "learning by doing", the organization builds up a reserve ofsuch assets under strong pressures. These invisible assets subsequently places the companyin good stead vis-a-vis the competition. [ITAMI87, p 161]

Geus asserts that "the root source of all competitive advantage is a company's ability to learn."[AGEUS88] Nonaka also reiterates that "the one sure source of competitive advantage isknowledge." Ray Stata [STATA92] also believes that organizational learning woulddistinguish successful companies in the future.

The essential point is that organizational learning adds unique values to the firm'sinfrastructure, products, and services. Learning also increases the firm's capacity for action byenhancing management cognitive capability and creating a broader inventory of employeeskills. In a networked organization, which has to respond to the business environment byrapidly re-configuring, organizational learning provides the basis for continuously re-shapingthe management's mental models to meet unpredictable conditions.

2.4 Learning Processes and Disciplines

2.4.1 Learning Loops

Chris Argyris and Donald Schon [ARGYRIS77] describe two forms of learning in relation tothe process of detecting and correcting error in organizations. They call these single anddouble loop learning.

Joop Swieringa and Andre Wierdsma [WIERDSMA92, p 13 & PP 35-43] further propose athird and more radical level, triple loop learning. According to them, organizations areguided by rules (what the members must do and allowed to do); insights (what the membersknow and understand); and principles (what the members are or wish to be). Rules, insightsand principles collectively define organizational behavior.

Single loop learning (also known as adaptive or lower-level learning) is straightforward. It isthe response to a condition that is out of alignment with present policies or objectives. Theresult of single loop learning is improvement. For example, a business process that is beingautomated in a way that parallels its old procedures incurs single loop learning or solution.The essence is to do more of the same, but better. Single loop learning tackles the "how"-questions while sidestepping the "whys". Rules are improved and solutions are sought withinthe existing insights and principles.

Double loop learning (also known as generative or higher-level learning), on the other hand,involves not only detecting the error but also inquiry about the underlying policies and goals.This form of learning is much harder as there is a natural tendency to avoid conflictingsituations if the organizational culture is ambivalent in its approach to handling errors. Theemployee may choose to "hide errors" rather than risk upsetting the status quo. Suppose, inthe process of computerization, the business process is radically changed to take advantage ofthe new opportunities afforded by the technology, the attempt to create organizational changerequires double loop learning. Double loop learning tackles the "why"-questions and leads torenewal of insights within the existing principles. Double loop solutions are usuallyattempted when single loop solutions fail to solve a problem.

Frequently, before double loop learning can take place, the organization has to go through aprocess of unlearning to free itself from the limitations imposed unconsciously as a result ofpast successes or failures. By developing a larger perspective and an awareness of newoptions, the organization begins to see opportunities for innovative learning , particularly onissues that may be so unique that there is no known solution and where trial-and-errorapproaches are not possible.

Single loop learning occurs at the operational level - changing rules or methods of doingthings to improve the performance of a particular system. New rules may result. On theother hand, double loop learning takes place at the conceptual level. This form of learninginvolves changing one's mental models and questioning the underlying assumptions to gainnew insights.

Triple loop learning applies to changes in the communally shared principles on which theorganization is based. The basic issues relate to the kind of organization the members wish itto be, the contributions they want to make and the values they consider important. Forexample, the chief information officer and some of the key information users may want toponder the enlarged role of IT and their mutual relationships. The result of triple looplearning is the development of new principles with which the organization can proceed to asubsequent phase.

Figure 2.1 shows how the three learning loops may be formalized [WIERDSMA92]

INDIVIDUALLEARNING

PRINCIPLES -- p- INSIGHTS -+ RULES--* BEHAVIORS-'- RESULTS

t SINGLE LOOP

DOUBLE LOOP

TRIPLE LOOP

Figure 2.1: Learning Loops

2.4.2 Learning Disciplines

Peter Senge has identified a number of learning disciplines -developmental paths foracquiring certain skills and competencies - that the individual and team should be aware of inorder to enhance learning capabilities in an organization. These disciplines, while usable inany organization, would be most potent in the knowledge-based organization that places apremium on learning.

The first discipline is personal mastery. Grounded in but yet going beyond competence andskills, personal mastery embodies personal vision and facing current reality. Personal visionenables one to focus on ultimate intrinsic desires as opposed to secondary goals. In contrast topurpose (e.g. excellence), which implies a general direction, vision spells out a concrete,specific destination (e.g. inventing a super-personal digital assistant) Senge refers to the gapbetween vision and current reality as creative tension. Creative tension can be resolved eitherby lowering one's vision or by bringing reality into line with the vision. Mastery of creativetension requires not the first, easier and ultimately self-defeating route but the second path -turning current reality into an ally, for example, by turning failure into an opportunity forlearning.

People who have achieve high levels of person mastery naturally integrates reason withintuition, and experience increasing connectedness with others. Organizations which fosterpersonal mastery value personal growth by encouraging personal vision, commitment totruth, and a willingness to face the gaps between the two. The leader's role is to be a modelthrough action - commitment to his own personal mastery.

The second discipline involves building and managing effective mental models that canaccelerate learning. It requires unfreezing outdated worldviews and replacing them with newmodels created through reflection and inquiry. According to Argyris, skills of reflectionconcern slowing down one's thinking processes in order to be more aware of one's mentalmodels and the way they influence one's actions. Inquiry skills concern one's face-to-faceinteractions with others, especially when dealing with complex and conflictual issues.

Reflection starts by recognizing "leaps of abstraction" which happens when one generalizeswithout testing. Such generalizations impede learning. Most managers are trained to beadvocates. Advocacy skills are useful in debating forcefully and influencing others but mustbe tempered by inquiry skills in order to promote collaborative learning. Finally, changes inone's mental models result in changes in actions. Here it is vital to recognize the gap between

one's espoused theories (what one says) and one's "theories-in-use" (the theories that laybehind one's actions) in order for learning to occur.

The-third discipline of shared vision provides the focus and energy for learning. Sharedvisions emerge from shared personal vision and are underpinned by a common caring, adesire to be connected in an important undertaking. Hence, personal mastery is thefoundation for developing shared visions. To enable sharing, visions must be articulated andlistened to. Visions are sustained by the enrollment, commitment and genuine compliance ofthe individual members who appreciate the benefits and do whatever it takes to fulfill thegoals.

In a corporation, shared visions enables the employees to have a common identity and putaside mutual mistrust. Shared visions are a first step to organizational learning. The loftytarget compels new ways of thinking and acting. Shared visions provide the incentive torecognize personal shortcomings, take risks and experiment.

The fourth discipline is team learning the process of aligning and developing the capacity of ateam to create the results its members truly desire. The discipline involves mastering thepractices of dialogue and discussion - two complementary skills. In dialogue, there is creativeexploration of complex issues while in discussion there is a search for the best view to supportdecisions. Potential conflicts and impediments to productive efforts such as defensivereasoning must be creatively handled.

Corporations today are very dependent on teams whether they are management teams, cross-functional task forces or otherwise. Mastering team learning benefits the organization in threeways. Team learning unlocks the intelligence and knowledge of many minds to thinkinsightfully on complex issues. Secondly, team learning inspires innovative, coordinated andspontaneous actions when the members act in ways that complement each other's actions.Thirdly, the practices and skills of senior teams can be passed to newer teams.

Systems thinking, the fifth discipline, integrates the other four disciplines, fusing them into acoherent body of theory and practice. Business organizations are systems that exhibitinterrelated behavior and actions that may take years to play out their full effects. Systemsthinking is a conceptual framework, a body of knowledge and tools that has been developedover the last fifty years, to make the interrelationships clearer and to evaluate changeseffectively. Two key principles of systems thinking are: "structure influences behavior" and"policy resistance" (the tendency of complex systems to resist change). Systems thinkingenables leverage - seeing where actions and changes in structures can lead to significant,enduring improvements.

The five disciplines should develop as an ensemble. Systems thinking links vision to theunderlying forces that must be mastered to get from the here to there. When developing newmental models, systems thinking provide the framework to examine these modelssystematically for possible flaws. These new models are then likely to recognize longer-termpatterns of change and the underlying structures producing these patterns. Systems thinkingalso provide a common language to teams for describing complexity and integrating theirindividual mental models. Team learning is expedited when the members use the samelanguage to articulate their views. Finally, systems thinking is pivotal in personal mastery byclarifying connectedess and commitment to the whole. System thinking leads to experiencinginterconnectedness of life and seeing wholes rather than parts.

It should be noted that the disciplines of building shared vision and team learning differ fromthe other three in that they are inherently collective in nature. The subtle differences between"interconnectedness" (system thinking) and "connectedness" (personal mastery) must be

noted. So, too, is the distinction between "commonality of purpose" (shared vision) and"alignment" (team learning). The former has to do with common direction and reason forbeing; the latter has to do with "functioning as a whole".

In a learning organization, direction setting is not necessarily a top-down process, althoughtop management must nurture a culture for shared visions to emerge from the lower levels.Merging thinking and acting at all levels is made possible by personal mastery and teamlearning. The nature of thinking is systemic as opposed to atomistic. Systems thinking skillsare needed to develop enterprise-wide awareness and improve mental models. Conflictresolution is achieved through dialogue and integration of diverse views. This calls for theneed to build shared mental models. The role of leadership in a learning organization isdirected at building shared vision, empowering workers and inspiring commitment.

Nonaka [NONAKA91, p1041 observes:...a company's vision needs to be open-ended, susceptible to a variety of different andeven conflicting interpretations...A more equivocal vision gives the employees andwork groups the freedom and autonomy to set their own goals. This is importantbecause while the ideals of senior management are important, on their own they arenot enough. The best that top management can do is to clear away any obstacles andprepare the ground for self-organizing groups or teams. Then, it is up to the teams tofigure out what the ideals of the top mean in reality.

2.5 Organizational Learning in practice

2.5.1 Planning as Learning

Geus has put forward the case for viewing planning as a learning process [AGEUS88].Strategic group planning at Royal Dutch/Shell takes the form of scenarios, one for everyconceivable crisis. As triggers for organizational learning, such scenarios, which may initiallybe obvious or far-fetched, eventually unleash a torrent of discussion after allowing forparticipants' mental models to absorb and re-adapt to the simulated events. In the next phaseof the learning process, operating executives attempt to draw conclusions from their revisedmental models of the business. These conclusions are then tested against experience. Eachstep in this cycle of perception, digestion, confirmation and action takes its own time whichvaries with each individual.

Geus contends that the only relevant learning in a company is the learning done by thosepeople who have the power to act. So the real purpose of effective planning is not to makeplans but to change the microcosm, the mental models that these decision makers carry intheir heads. Organizational learning begins with the calibration of these existing mentalmodels. However, eventually, the users (planners) have to integrate the new knowledge intotheir working mental models, and translate the new knowledge into positive action.

2.5.2 Learning in R&D Processes

Meyer's study [MEYER92, pp. 172-173] of some successful learning organizations reports thatlearning in R&D processes takes place in most cases through the planning, evaluation andcontrol processes. Planning is used as a way of exchanging ideas between laboratories,advisory and strategy groups. Communication and cooperation are emphasized. The controlprocedures, mainly in the form of standardized progress reports, are diffused efficiently totargeted R&D and marketing staff. The less successful organizations tend to raise planning tothe highest level, leaving out the managers who would eventually be executing the tasks.

In cross-boundary R&D, Meyer argues that it is crucial to have common scenario building,common preparation of the strategy formulation and shared projects. Organizational learningshould be accorded priority over planning efficiency.

2.5.3 Total Quality Management

Total quality management (TQM) provides yet another setting for organizational learningacross multiple levels and functions. TQM results are derived from skill which is gainedthrough serious application of TQM methods, mutual learning, and openness to learning.According to Shoji Shiba, Visiting Professor at MIT and authority in TQM [SHIBA93]:

TQM cannot be practiced effectively with knowledge only. Understanding and skillare necessary, and these start with aggressive learning and listening...Whenimprovement activities are being presented to management,...an aggressivelylistening manager will have a question or insightful observation. Simply listeningpassively... is not aggressive learning.

Mutual learning takes place in work groups in which no one is much more knowledgeablethan the others. Openness to learning is possible only if participants refrain from three basicsyndromes : (1) "I already know it", an excuse to tune out information and abandon theprocess of acquiring skills; (2) "Not invented here", rejecting ideas simply because they arenew or foreign; and (3) "Prove it to me", an insistence on accepting only objective, codifiedknowledge and rejecting tacit, subjective knowledge.

While TQM provides strong support for personal mastery, mental models and team learning,it is less so in shared vision building. The scope of some TQM projects may be sufficientlybroadly defined to provide room for effective systems thinking, e.g. rebuilding a damagedcustomer relationship. Analog Devices' experience shows that TQM can be effective inaccelerating organizational learning.

The themes of TQM - management by fact, management by process, focus on a vital few,systematic problem solving and continuous improvement, cross-functional management andcommitment to shared learning - are convergent with the goals of organizational learning.However, it is also pointed out that much of learning in TQM really takes place at theoperational level. Beyond the initial mental breakthrough, there is little motivation forlearning at the conceptual level. A manager can go on advocating improvements within thecurrent framework of the organization's policies and traditions without gaining much insightabout the whole system with which to reframe problems in a totally different context. [KIM90]

It has been observed that the near-religious fervor that TQM methodologies are pursued oftenleaves scientifically-oriented individuals uncomfortable [LEVY93]. Another reason is theinability or unwillingness to face up to organizational weaknesses. The failures of TQM inmany organizations have probably some roots in such resistance. While there may be ways tointegrate learning effectively into the TQM methodologies, one can argue that learning (perSenge and Argyris) is a more subtle approach to organizational improvement over the longrun and thus more likely to succeed in individual-oriented cultures. Furthermore, it must benoted that in most organizations in which TQM is successfully practiced, top management hasto take an active lead - a top-down process. Senge's notions of organizational learningrequires across-the-board participation from the lower levels in setting personal visions whiletop management plays a nurturing role.

2.5.4 Modeling as Learning

Drawing from the experience of companies like Hanover Insurance, Senge and John Sterman[STERMAN91] propose the creation of learning laboratories or "microcosms" of real business

settings where managers play roles in a simulated environment. The main purpose is toaccelerate learning by mapping (explicating and structuring assumptions),challenging(revealing inconsistencies), and improving (continually extending and testing)mental models.

In such a learning laboratory, participants discuss the issues that are of concern to them. Theythen use causal diagrams to map their mental models. Using computer simulation tools basedon systems dynamics concepts, these managers next formulate their strategies and test theirexpectations against the computer's results. In the process of reflecting on the discrepancies,the managers revise their mental models for the next set of experiments.

"Modeling as learning" aims to adapt mental models with the ultimate goal of improving theorganization. Ideas are expressed in an explicit, concise way. Systemic complexity andpersistent problems that defy efforts at improvement are revealed. It is claimed that learningand intuition are both created through such risk-free experimentation. Senge and Isaacs arguethat such computer-based learning environments (CBLE) can support double loop learning[ISAACS92].

The major challenge of this approach to learning is to devise processes whereby managers canmove continually between the virtual world of the learning laboratory and the real world ofmanagement practice. Insights gained from CBLE experiments must somehow be transferredto the workplace. For a model to produce reliable predictions of real world systems, it mustbe a precise and complete representation of reality - an unattainable goal. An oversimplified,"academic" model will not be taken seriously by managers or executives. There are otherpitfalls, for example, CBLE's may be misused to prove a point or validate a predeterminedconclusion. There is also a tendency to treat CBLE's results as an oracle.

2.5.5 Benchmarking

Robert C. Camp, a leading advocate of benchmarking, defines it as " the search for industrybest practices that lead to superior performance." Benchmarking's popularity is partly to theMalcolm Baldrige National Quality Award which requires all entries to be benchmarked.Global competition is another major driver as corporations seek out the best practices tobecome the best. In benchmarking, the organization has to ask what and how it can learnfrom internal sources and from other companies.

Wick and Leon [WICK93] discuss the benchmarking process as a series of steps: (1)brainstorming what needs to be learned; (2) identifying the best company (not necessarilyfrom the same industry) in the field; (3) discovering everything about the company to bebenchmarked while keeping specific goals in mind; (4) evaluating the performance gap andhow to benefit from what has been discovered; (5) developing and implementing action stepsto close the gap; and (6) putting the new knowledge to use.

Camp argues that the basic willingness to share non-proprietary and non-confidentialinformation is based on a mutual desire to uncover and understand the industry bestpractices. The new-found knowledge is the cumulation of years of experience by thebenchmarked company. However, it is doubtful that the expertise thus acquired can be putinto use without going through a double loop or generative learning process.

2.5.6 Experiential Learning

It is also useful to differentiate between learning by doing or experiential learning (imitatingand attempting) and learning through adapting one's mental models. In the former,knowledge is embedded in practical action that is evanescent. This action-centeredknowledge is not recorded; it can be lived and witnessed but it vanishes into mere potential

when the action is completed [ZUBOFF84,p175]. Building mental models, however, isbelieved to have a lasting effect on the learner, especially if opportunities exist for continuouslearning. The knowledge gained will be warehoused for future use. In Shell's planning andTQM processes, both forms of learning are involved; planning emphasizes more on mentalmodel building whereas TQM has more of action-centered learning.

Hence, in order for learning to take place, organizational processes should be conscientiouslyplanned with clear learning goals in mind . The eventual implementers must be involved anda common understanding must be reached among all the parties. The goal is to developshared mental models amongst the participants. Learning is effective when there isknowledge-sharing, openness and objectivity. The most effective way to learn - introduceknowledge and modify behavior - is by working with small groups that have the power andresources to enact change [STATA89].

Nonaka [NONAKA91, p104] also highlights the role of teams in successful learning:Teams play a central role in the knowledge-creating company because they provide ashared context where individuals can interact with each other and engage in constantdialogue on which effective reflective depands. Team members create new points ofview through dialogue and discussion. They pool their information and examine itfrom various angles. Eventually, they integrate their diverse individual perspectivesinto a new collective perspective.

2.6 Learning Disabilities and Impediments

Schein and Senge have identified a number of learning constraints and disabilities at both theindividual as well as organizational levels.

One basic obstacle is the lack of personal motivation. Some individuals are unable to see theneed for learning and lack the confidence that learning is possible. This problems surfaces inthe form of anxiety and defensive denial. For example, production operators who only focusonly on their tasks on hand have little sense of responsibility for the results produced when allpositions interact. The tendency then is to finger-point at others when problems occur. Theremedy is to encourage "personal mastery" or "connectedness".

Individuals face cognitive limitations when attempting to process large amounts ofinformation, and in observing and drawing inferences on complex relationships. Cognitivelimitations may be partially overcome by training in inquiry methods and application ofsystem dynamics. Group-centered problem-solving such as quality circles and other totalquality management techniques also help to develop cross-functional analytical skills for thelower-level workers. A caveat is that an over-fixation with immediate events can also blindthe organization to slow, gradual processes that are ultimately destructive. Such slowprocesses may include environmental decay, decline in educational levels of new hires, andobsolescence of process technology.

Work teams may also encounter interpersonal limitations and inability to maintain dialoguewith others, especially across group boundaries. This results in the inability to build andsustain collaborative team relationships with internal fellow workers as well as externalpartners. The remedies are training in dialogue and group dynamics. However, theorganization has to watch out for appearances of cohesiveness in its management team as aguise for "skilled incompetence" whereby serious personal reservations are stymied and joint-decisions are merely watered-down compromises.

Cultural limitations must be addressed. These relate to shared tacit assumptions that favorthe status quo based upon nation, ethnic group, total organization, occupational community,

functional group, group membership or corporate norms. Examples of cultural limitationsmay be recent but not well-founded organizational "philosophies" like fostering internalcompetition, and organizational downsizing, when practiced without an appreciation of theirgoals. The remedy is to educate employees to analyze and manage cultural assumptions.

The organization's incentive and reward system directly influence the will to learn. A systemthat rewards employees who carry out their work in a routine manner over those whodiscover and attempt to fix errors will ultimately destroy the will to improve. The result is alearning disability. On the other hand, an incentive system that rewards innovativeapproaches to the organization's problems will encourage employees to experiment with newsolutions.

Chris Argyris [ARGYRIS91] points out, however, that it is frequently the smart, well-qualified, and high-performing members of the organization who have the most problems inlearning. These professionals are willing participants to external organizational factors likejob redesign, compensation programs and leadership training, but are resistant to effortsdirected at their own continuous improvement. They do not resist change; they resist beingchanged. More than just an issue of motivation, learning requires breaking old habits ofdefensive reasoning and replacing these with critical self-analysis. Employees have to beaware of how their behavior can adversely impact the rest of the organization.

Hence, an organization's learning disabilities may arise from:1) not perceiving the need to learn because narrow work functions result in localized problemsolving;2) lacking the courage to learn, a situation covered up by defensive denial and blaming otherssuch as "the system";3) lacking the will to learn due to the fear of change and the absence of a reward system thatencourages learning;4) lacking the ability to learn due to the bureaucratic culture that separates "thinkers" from"doers".

2.7 Impacts on Networked Organizations

Learning cannot and should not be limited to a group or scattered groups within anorganization. In many global businesses today, learning spans national boundaries. Learningabout different markets, different problem-solving methods, different sources of technologicalprogress, different cultures, different competitors is definitely enhanced by creating aninternational network. For example, internationalization of R&D, despite the risks ofdisseminating proprietary know-how to other countries, enables the rapid diffusion oflearning throughout the organization.

In the words of C.A. Bartlett [BARTLETT86]:

In the integrated network configuration, national units are no longer viewed only asthe end of a delivery pipeline for company products, or as implementers of centrallydefined strategies, or even as local adapters and modifiers of corporate approaches.Rather, the assumption behind this configuration is that management should considereach of the worldwide units as sources of ideas, skills, capabilities, and knowledgethat can be harnessed for the total benefit of the organization.

And Shiba follows with:

No significant management innovation could ever be achieved by one companyworking in isolation. The complexity of the issues is vast. The range of experiments

you need to run is tremendous. The discipline required to run those experiments andstudy them rigorously, is much beyond even the largest corporation.

Bartlett and Goshal [BGOSHAL89] identify the three critical management tasks in themanagement of transnational corporations as efficiency, local responsiveness, and worldwidelearning and knowledge transfer. Learning is the source of core competencies for theorganization, enabling the coordination of diverse production skills and integration ofmultiple streams of technologies. As Prahalad and Hamel [PRAHALAD90] point out, theskills that together constitute core competence must coalesce around individuals whoseefforts are not so narrowly focused that they cannot recognize the opportunities for blendingtheir functional expertise with those of others in new and interesting ways. Organizationallearning provides the context and motivation for individual learning to be crystallized in theform of organizational change and improvement.

At ABB, individual profit centers, plants and companies within a "country structure" interactswith Business Areas (hydropower plants, complete rail systems, etc.), creating a fluid,internally competitive global network. Sune Karlsson, Chief of Power Transformers BusinessArea, says that they are "not a global business" but "...a collection of local businesses withintense global coordination." Besides the "hard" advantages from scale economies, there is the"soft" advantage. ABB has people around the world "working on the same problems andopportunities day after day...and learning a tremendous amount." In order to create "aprocess of continuous expertise transfer", Karlsson has created "forums for exchange."[PETERS92]

There are strategy meetings four to six times a year between Business Area board members.Staff with special responsibility e.g. purchasing, travel constantly to confer with countrymanagers, pushing the overall learning and coordination agenda forward. Functionalcoordination teams, made up of expert members from various operations, meet a couple oftime a year. Karlsson admits that "sharing of expertise does not happen automatically...peopleneed to spend time together, to get to know and understand each other."

2.8 Discussion

The networked corporation and the knowledge-based organization overlap in their structuresand cultures. Both will thrive as learning organizations. A networked organization has to beincreasingly knowledge-based. At the same time, organizational learning will facilitateperformance improvement in, and the continuing renewal and development of, theknowledge-based networked organization.

Learning organizations therefore are typically characterized by network structures whichcoordinate through dialogue. The culture is likely to be flexible and problem oriented withample room for members to express their creativity. The learning organization also needs tobe supported by an information infrastructure that provides for reflection and action.

Being alert to errors and their detection is a learning process. By switching from single anddouble loop learning as circumstances require, the organization is able to proceed to the nextphase of renewal. Learning disciplines have to be inculcated in the workplace andcontinuously practiced. The five disciplines of systems thinking, team learning, shared vision,building mental models and personal mastery will enrich both the individual and theworkgroup. The organization also has to be aware of learning disabilities and ways toovercome them. Ultimately, the organization's goals are achieved, not through topmanagement direction, but through commitment to shared visions by all the employees.

Chapter 3: Organizational Learning in IT Processes

3.1 An Overview of IT Issues and Challenges

Information technology has become an indispensable tool of business. Yet the recurring theme innumerous management seminars and publications relate to the need to derive more benefits fromIT. For example, in a recent Computerworld/Andersen Consulting survey of 203 chief executive,chief operating and financial officers, less than half the respondents believe they are getting theirmoney's worth out of their information systems (IS) investments. [COMWORLD93]

One aspect of the dissatisfaction relate to the perception that IT has not contributed up tomanagement's expectations to the organization's competitive advantage and overall productivity.There is a vague conviction that IS should be doing more to help the bottom line and grow thebusiness. A more pressing problem is the inability of IT solutions to keep up with organizationalchanges. The IS function appears to be failing to deliver solutions in a timely manner and in aform that the users expect. There is also a perception that the IS function is not learning - notadapting to changes and providing creative solutions to the organization's problems. Othercritics also point to the tendency of the IS function to keep a tight rein on IT resources.

Many companies are also beginning to use new yardsticks for measuring the effectiveness of ITinvestments. Beyond the traditional cost and performance methods, today's IS functions are alsoassessed on such key areas as customer satisfaction, user-friendliness, return on investment, dataaccuracy and information exchange among departments. There is a growing sense that IS shouldbe judged as part of the whole project and organization rather than strictly on its own terms. Inparticular, IS is expected to take a strong participative role in re-engineering or business processredesign. [COMWORLD93]

Based on "The Management in the 1990s Research Program" at MIT during the 1980s, Michael S.Morton identified the following strategic implications of IT [CORP90]:1) IT is enabling fundamental changes in the way work is done.IT can radically alter the basic cost structure of a wide variety of jobs, provided there ismanagement leadership and employee participation. IT can have significant impacts onproduction, coordination and management activities.2) IT is enabling the integration of business functions at all levels within and betweenorganizations.Integration shows up within the value chain, at end-to-end links of value chains betweenorganizations, in value chain substitution via subcontract or alliance, and in the creation ofelectronic markets.3) IT is causing shifts in the competitive climate in many industries.IT is interrelated with the economic forces of competition. As the traditional economic forces(such as competitor actions) change, or as the economics and functionality change, the position ofthe organization relative to its competitors will change.4) IT presents new strategic opportunities for organizations that reassess their missions andoperations.Organizations respond to their environments by going through the stages of automating,informating and transforming. Going beyond automation of production and services, aninformated organization requires new ways of thinking, creates new information-based by-products (new knowledge) and induces the development of new skills and information to theextent that new market opportunities open up.5) Successful application of IT will require changes in management and organizational structure.The traditional hierarchical and matrix forms will have to give way to networked structures.Horizontal and vertical working patterns can be created as required. This increasingly popular

option leverages IT's ability to affect coordination by shrinking time and distance to improve theorganization's responsiveness to market forces.

The above implications suggest that the traditional perception of IT as a data-processing tool isinadequate. In addition to the strategic implications, IT should also be viewed as a strategicresource that should be managed interdependently with other key organizational resources.Moreover, IT processes impact all level of the organization and re-vitalizes routine tasks, thuscreating numerous learning opportunities in all the application domains.

3.2 Shared Responsibility for IT

It is pertinent to address the issue of responsibility for IT in a networked organization. Boyntonet al [BOYNTON92] argue that the best way to link IT consistently to a firm's day-to-day, corebusiness processes is to carefully distribute IT management responsibilities to line managers.

The rationale is twofold. Firstly, overdependence on the central IS functions leads to resistance tothe effective use of IT as the line managers may not want to rely on resources that they neithercontrol nor fully understand. With shared, dispersed responsibility, line managers will use ITresources more effectively, learning to apply IT to business tasks just as they apply human,financial, and other key resources to business opportunities, problems and threats. Secondly, bysynergistically uniting the IT function and business knowledge, the line manager is able to makestrategic decisions that are driven by day-to-day management.

Various other studies have also emphasized the pre-eminence of action-oriented skills ofexecutive managers. For example, Daniel Isenberg's research on "how senior managers think"found that top managers think in ways that are highly "intuitive" and integrated with action[ISENBERG84]. Says Isenberg : "Thinking is inextricably tied to action...Managers developthought about their companies and organizations not by analyzing a problematic situation andthen acting, but by thinking and acting in close concert." Zuboff also supports this view thatlearning and influence of top managers occurs primarily through action [ZUBOFF84,p106].

Over the years, line managers have increasingly gained IT management responsibilities. ITresources have been brought to bear upon business and strategic challenges associated withcross-functional integration, coordination and control of mutually dependent value chainactivities and team development across organizational and geographic boundaries.

Line and IT managers should therefore team up to form an "IT management architecture" or IT-line management partnership. Such an architecture will enable IT decision-making strategies toalign with business strategies. At the same time, it will defuse criticisms directed against the ITmanagers for "hoarding information, accumulating power, and trying to keep others from havinga finger in the pie." [PETERS93]

Rockart summarizes this trend:

Information technology has become inextricably intertwined with the business. It has,therefore, become the province not only of information systems professionals, but ofevery manager in the business no matter what his or her level. [ROCKART86].

Hence, in order to enable organizational learning, top and line managers need to become activeparticipants in key processes within the organization. The almost universal acceptance of IT as afixture in any modern corporation makes IT a most suitable vehicle for action-centeredorganizational learning. By being involved in the IT planning and implementation, IT, line and,to some extent, top management participate in team learning processes. In so doing, they

develop a shared vision and facilitate the integration of systemic solutions into the organization'sinformation infrastructure.

3.3 Information Politics

Thomas Davenport et al [DAVENPORT92] argue that efforts in many companies to createinformation-based organizations have led to failure primarily because management neglected ormismanaged information politics. The paper identified five models of information politics : (1)technocratic utopianism, (2) anarchy, (3) feudalism, (4) monarchy, and (5) federalism. Of the five,only monarchy and federalism are viable in any type of organizations over the long run. Mostcompanies are likely to have information politics that are hybrids of two or more of these forms.

In a monarchy, the CEO, or someone appointed by him, dictates the rules for how informationshould be managed. Power is centralized, and departments and divisions have substantially lessautonomy regarding information policies. This top-down model may be appropriate for firmsthat have difficulty achieving consensus across business units. The shortcomings of themonarchy approach are the lack of continuity of policies as power changes hands and the relativewillingness to share information. The celebrated case of Ms Fields Cookies' IT infrastructure is anexample of information monarchy. [FIELDS]

In contrast, federalism is distinguished by the use of negotiation to bring potentially competingand non cooperating parties together. The approach to information management is based onconsensus and negotiation on the organization's key information elements and reportingstructures.

Firms that adopt the federalism model typically have a culture that encourages cooperation andlearning. With the CEO's support and the other divisional heads, the information manager is ableto create a shared information vision. The authors argue that it is apt to be the most effectivemodel for companies that rely on individual initiative for generating collective action. Thefederalist approach supports both autonomy and coordination. Federalism is therefore the mostsuitable model for networked organizations. It also provides the best environment for learning totake place.

3.4 The Organization as a Learning Laboratory

Other studies have put forward the notions of R&D organizations as learning systems[CARLSON76] and the factory as a learning laboratory [LBARTON92].

Dorothy Leonard-Barton views a learning laboratory as an organization dedicated to knowledgecreation, collection and control. In a learning laboratory, vast amounts of knowledge and skill areembedded in physical equipment and processes and embodied in people. Even more importantare the managerial practices and underlying values that constantly renew and support theknowledge bases. Learning requires creation of both external and internal knowledge for bothcurrent and future operations.

Applying these ideas to a factory, Leonard-Barton says that a factory that functions as a learninglaboratory is totally integrated. In making a study of Chaparral Steel, Leonard-Barton notes thatthe management emphasizes homogenizing the level of knowledge by sharing. Even leadoperators are engaged in transmitting knowledge flows to their peers. Knowledge-sharingenables a practice called "vice-ing" whereby a foreman from another shift is temporarilypromoted to take over in his peer's absence. While hierarchical boundaries are minimal, thetraditional horizontal boundaries are redrawn. For example, production workers handle 40

percent of maintenance tasks and everyone in the marketing department is considered asalesperson.

The culture is oriented toward egalitarianism, respect for the individual, continuous educationand experimentation, knowledge sharing and consistent mental model building across the ranks.There is no R&D department as "everyone is in research and development...the plant is thelaboratory". However, for economic reasons and to ensure that the company does not miss outon key technologies, there is also extensive networking and alliances.

Paradoxically, Leonard-Barton concludes, the system's interdependence is also a potentialweakness as the learning laboratory would have difficulties recreating its unique qualitieselsewhere.

3.5 Knowledge Creation in IT Processes

Analogously, one can hypothesize the networked organization as a learning system. Thebusiness processes would have to be able to address cross-boundary needs while giving ampleroom for innovation and knowledge creation. As a major driver of organizational change,information technology processes provide broad opportunities for systems thinking and henceknowledge creation.

Four types of knowledge-creation and collection activities are distinguishable in IT planning,implementation and usage:(1) problem solving (designing and implementing an IT architecture);(2) internal knowledge integration (across functions and projects);(3) innovation and experimentation (to mix and match numerous technologies and build for thefuture); and(4) integration of external information (from suppliers, customers, partners).

Much has been done to systematize IT planning, implementation and usage. However, given theever increasing complexity of organizations, particularly of the networked variety, a great deal ofmanagerial and professional judgment is needed to execute these well-defined methodologiesboth because of the maze of conflicting organizational requirements as well as the spectrum ofavailable technologies. Furthermore, information technology advances at a pace that rapidlyobsolesces planning.

In considering IT planning, implementation and usage as learning processes, we have to view theentire organization, networked or otherwise, as a learning laboratory. IT has to be viewed as anequal partner with the rest of the organization's business activities. The recognition of thisinterdependence enables the users to form a shared vision with the IT function in a way thatenhances mutual learning.

Ray Stata [STATA89] of Analog Devices, Inc., sums it up:

Management information systems transform data into information and then helpmanagers transform information into knowledge and knowledge into action. Thechallenge is deciding what information and knowledge - in what form - are needed. Ifwe keep organizational learning in mind as a goal of information systems design, thenwe are more likely to generate the information and knowledge that managers need totake effective action.

3.6 A Cyclic Learning Model

Barbara Carlsson et al [CARLSSON76] use a four-step repetitive cycle as a framework fororganizational learning process in R&D organizations. The basis of this framework is D.A Kolb'swork on individual experiential learning [KOLB71].

Essentially, Kolb maintains that individual learning is a cyclic process : doing -> reflecting ->

thinking -> deciding -> (re)doing. Kolb's learning process requires orientations that are polaropposites: active and reflective; concrete and abstract. The shifting orientation results in fourkinds of activity, each of which is required at some stage of the learning process. Kolb calls thisprocess problem-oriented learning as problems represent both the motive and vehicle forlearning. Some other writers describe it as "learning by doing" or experiential learning.

We shall adapt Kolb and Carlsson's learning models to IT-enabled organizational learning.Arguably, every non-trivial IT project will go through a cycle of reflecting (feasibility study),thinking (planning), deciding (implementation), and doing (usage). Each phase of the cycle callsfor additional knowledge creation and sharing. This knowledge then becomes an input to thenext stage. For example, implementing groupware generates new knowledge about itspotentials and shortfalls. Such knowledge will be useful when the organization is evaluating anew substitute technology.

Our model for IT-enabled organizational learning is presented in Figure 3.1. We shall call this theProblem-oriented cyclic learning model or simply cyclic learning model.

Concrete

Divergence

Convergence

Abstract

Active d ~Reflective

Figure 3.1: A Model for IT-centered Organizational Learning.

Each of the four phases in introducing IT into an organization -feasibility study, planning,implementation and usage - are characterized by a different set of activities. The nature of theproblems encountered at each stage too are different.

The activities are summarized below.

1) Divergence (concrete and reflective).This kind of activity is required to seek background information and sense opportunities,investigate new patterns of work, recognize shortfalls and problems in using the technology, andgenerate alternatives. There may be problems during this activity - excess alternatives, e.g. toomany competing technologies.

2) Assimilation (abstract and reflective).This kind of activity is required to develop the IT architecture, compare alternatives, establishcritical success factors, formulate plans and identify potential problems. One deficiency of thisactivity is that the plan may not be implementable.

3) Convergence (abstract and active).This type of activity is required to select among implementation alternatives, focus efforts, designnew processes, and make decisions. A possible problem here is misunderstanding the planninggoals.

4) Execution (concrete and active).This kind of activity is required to advocate position or ideas, set objectives, commit to schedules,commit resources, and adapt to new processes. The problems here may arise from a lack ofcongruency between the employee behavior and the organizational norms.

The four-step cycle can help a project team leader understand the changing characteristics ofinteractions among the members. For example, in some meetings they may find themselves opento ideas and value inputs from others while at other meetings they generally do the opposite.The reason is that the first condition prevails when the project is in a Divergence stage and thelatter occurs when the project reaches the Convergence or Execution stages.

In a learning system, the project manager's role should be viewed as one providing leadership inthe learning process. When the project is going smoothly, the manager's role should be one ortwo phases ahead of the team. For example, if the team is engaged in ideas generation, themanager will be thinking about the criteria for solution. In times of crisis, the manager shouldmove into the same stage as the project team. For example, the manager may work with the teamto explain an unexpected result when the team's progress is stalled. For the learning process toachieve results, the manager must pull the team around the cycle, not across it.

The Divergence and Assimilation stages provide opportunities for double-loop learning; theConvergence and Execution stages mainly support single-loop learning. During the Divergenceand Assimilation stages, the project team is free to examining critically all assumptions about thecurrent and future organizational policies and objectives. This rethinking allows the team topropose a new framework for exploiting IT in the organization. During the Convergence andExecution phases, the project team sets objectives for successful application of IT to supportpresent and newly-revised policies. Incremental improvements may be made during thesestages.

Of course, learning is not limited to one cycle. As the project team implements a new project, itrestarts the learning cycle. Each repetition reinforces the previous learning experience, addingnew insights on the organization's processes.

The attractive aspects of this model are manifold. First of all, learning is not equated withacquisition of knowledge. The central point is that what people have learned must be evidentfrom what they do. Secondly, in organizational learning, problems are both the stimulus and the

medium for learning. Thirdly, this model can also be applied to a broad spectrum of everydayorganizational activities. Everyday experience can be a component of the learning process, andthese often unconscious activity processes, in conjunction with thinking and deciding, can beraised to the level of conscious learning. Conscious learning is believed to lead to a higher levelof competence than unconscious learning. [WIERDSMA92]

While allowing that the most powerful form of learning comes from direct experience, Senge hasa word of caution on relying on experiential learning exclusively. Humans have a "learninghorizon", a breadth of vision of time and space within which the effectiveness of an experiencecan be assessed. When actions have consequences beyond the learning horizon, it becomesimpossible to learn from direct experience. Senge notes (SENGE90, p231:

Herein lies the core learning dilemma that confronts organziations : we learn best fromexperience but we never directly experience the consequences of many of our most importantdecisions. The most critical decisions made in organizations have systemwideconsequences that stretch over years or decades...Cycles are particularly hard to see, andthus learn from, if they last longer than a year or two.

The Problem-oriented learning model can be the basis of management decisions such as:1) Staffing. By understanding the requisite skills in each stage, management can assignemployees to take advantage of their abilities and to offset the learning disabilities of other teammembers.2) Reward systems. Organizational policies and reward systems can be used to support anappropriate balance of learning abilities. Organizations and professional disciplines oftendevelop values that favor activity in one phase over the other. Management can use the model tohelp integrate values with the needs of the organization.3) Remedial action. Specific problems can be identified, and strategies for remedial action may bebased on the model.

3.7 Discussion

IT processes are ideal candidates for enabling organizational learning for the following reasons:(1) their use pervades throughout the organization, (2) most levels of management have someprior experience with IT applications, (3) they enable knowledge creation and sharing, (4) theyare highly action-and-thinking-centered, and (5) its interdependence on the business strategiccontext continually creates new need for adaptations.

Learning should not be restricted to any one group or level of workers. Learning should bedispersed, between peers and across functional disciplines and ranks. Line and IT managershave to work out a mutually beneficial agenda in cooperating to form an action-centered learningteam in order to synergize their respective strengths to re-design business processes. Topmanagement should provide an environment in which IT-enabled organizational learning cantake place.

Summary of Part I

Networked organizations are expected to be the wave of the next decade. Severalorganization experts have pointed to the need to create a learning culture in order forsuch organizations to grow and be self-enhancing in terms of work, knowledge and inter-personal relationships. In other words, a networked organization should also be aknowledge-creating or learning organization.

The most effective learning process for management should be action-and- thinkingoriented. IT planning, implementation and usage are processes that offer some uniqueopportunities for line and IT managers to learn and collaborate in redesigning theorganization's information infrastructure for the future. In so doing, they are alsoredesigning the business processes.

The IT-line management partnership provides the following advantages[HENDERSON89.1]:

1) Mutual benefits. By viewing the IS function as a service organization providingsupport and resources to line management in their pursuit of business objectives, thebenefits may be identified as financial contribution (made by IS to the achievement ofbusiness objectives), operations efficiency (as a result of effective partnership indevelopment, implementation, etc) and quality of work life (as a result of improvedworking relationship).

2) Commitment. Members in the partnership share goals that lead to effective utilizationof IS services and hence to an effective working relationship.

3) Shared knowledge. The IT management architecture leads to mutual understanding oftechnology and business practices.

4) Mutual dependency on competencies and resources. Line and IS organizationstogether contribute critical skills, experience and physical assets necessary to managedata across the corporation.

This partnership fosters an environment for mutual learning.In planning and implementing IT projects, and in the subsequent utilization of ITresources, organizational knowledge is created. This knowledge is reinforced ifmanagement and the users are aware of the learning potential in these activities and settheir goals accordingly. Management and users have to learn to learn, i.e., intentionallyseek out learning opportunities. This goal may require unlearning past experiences thatnow inhibit new creative thinking and discarding defensive reasoning. Understandingthe nature of organizational learning through understanding learning processes,disciplines as well as the corresponding disabilities and ways to overcome them is thefirst step toward this goal. Figure S1.1 shows the arrows as bi-directional to highlight theinterdependence of the three components.

ORGRNIZRTIONALKNOWLEDGE

IT PROCESSES LEARNING- planning PROCESSES &- implementing DIRCEPIES- usageFigure S1.1: Linkages between organizational knowledge, learning

disciplines and IT processes.

Part IIWe shall next examine a number of key information technologies that may impactorganizational learning. These technologies are chosen mainly for their wide popularity andability to significantly impact the operations and decision-making process in networkedcorporations. The approach used to analyse these processes are also applicable to other ITprocesses, e.g., software development, and the results should also provide insights into otherIS applications.

Chapter Four focuses on electronic mail - specifically, as a tool for software messaging andcomputer conferencing - and examines its potential to support organizational learning. E-mail has been in use for several years as a communication medium. Its impact on work andculture has been tremendous, redefining team building across time and space and enablingwork that would otherwise be impossible. The limitations of e-mail as a learning tool will bediscussed.

Chapter Five examines a broader class of emerging technology, workgroup software orgroupware. Groupware enhances teamwork by facilitating changes to business interactionsand work practices. The successful implementation and use of groupware, however,depends on the organizational structure and culture. The focus of the discussion is onimplementation of workflow automation.

Finally, Chapter Six looks at a global view of IT strategic planning and its relation to theusers. Planning and implementation of open network environments, particularly in relationto the rightsizing of information resources require top management involvement andsignificantly influence the subsequent IS deployment and usage within the organization. Weshall consider two different models of information infrastructure planning andimplementation. By contrasting these two models and their respective approaches toplanning, we highlight some of the differences in organizational learning processes.

In each case, we shall look at how the activities generated in these IT processes with respectto : 1) the Problem-oriented learning cycle (or experiential learning); (2) the learning loops -single, double and triple loop learning/solution/planning; and (3) interactions with thelearning disciplines - personal mastery, team learning, mental models, shared vision andsystem thinking. In particular, we are interested in the learning opportunities, disabilitiesand remedies in each process.

In the final summary, we shall also make a subjective evaluation of the relative strengths ofthese IT processes in supporting the learning disciplines and processes. The usefulness ofsuch an evaluation lies in highlighting the relative intensity and extent of learning that ispossible under each set of circumstances.

Every information technology impacts the organization differently. Organizational learninglikewise will be affected to differing extents. Employees will learn more enthusiastically ifthey perceive an immediate return or benefit for their efforts. A project team that plans andimplements an information technology will definitely learn far more about organizationalissues than mere users. Not only do such project team members improve their competenceover the related technical and organizational issues, they also would have developed acommon mental model of the application areas over which the technology is used.

Planning requires far more knowledge creation and sharing than using a software tool like e-mail. In general, a process that requires innovation and creative thinking should providebetter learning potential. Nevertheless, attempting to learn from IT processes will inevitablyrun into impediments as a result of cultural, interpersonal and cognitive limitations. Thesedisabilities will need to be overcome for effective learning to take place.

Chapter 4: Using E-mail

4.1 Overview of Electronic Mail

The Electronic Mail Association defines e-mail as follows:

Electronic mail is the generic name for non interactive communication of text, data, imageand voice messages between a sender and designated recipients by systems utilizingtelecommunications links.

The definition of electronic mail (e-mail) rules out the telephone, which is interactive. However,electronic document interchange (EDI) systems between organizations are included. E-mailtherefore creates a new form of communication - store and forward electronic communicationwhose content is typically unstructured. On the other hand, the format for addressing themessage and its characteristics (urgent, broadcast, restricted to, etc.) are usually structured.

While e-mail is mostly a stand-alone capability in a user workstation, EDI is integrated withexisting data-processing capabilities of the organization. Other variants of e-mail include theelectronic bulletin board which arranges incoming messages in chronological order for specialinterest groups, and computer conferencing systems which sort messages by topic. However,with the advent of more sophisticated groupware, several of these capabilities are bundledtogether.

Today's e-mail systems encompasses private systems purchased from the major computervendors, public messaging systems provided by VANS (value-added network suppliers), andLAN (local area network) based mail systems primarily for personal computer users. However,in the short term, the greatest problem facing the networked organization is the lack ofconnectivity between e-mail systems. These woes include the inability to link different packagesacross heterogeneous networks and the problem of how to synchronize distributed directories ofnetwork and mail users. the However, mail standards (ie., APIs, file formats, transports, anddirectory services) are gradually emerging.

Internet, a global communication network started by the US Department of Defense, has reachedout to private and public organizations including academic institutions in developed as well asthird world countries. In the corporate world, IBM's private network is believed to be the mostextensive.

According to a study by The Yankee Group, the number of LAN e-mail users is expected to growfrom 15.1 million in 1993 to 38 million by 1995. The overall total number of messages transmittedwithin the Fortune 2000 firms in North America is projected to grow from 6.1 billion in 1993 to14.5 billion in 1995. Hence, not too surprisingly, numerous companies have sprung up to providee-mail products and services.

E-mail may be thus used to communicate within a work group, across an enterprise, or externallywithin customers, suppliers, partners or even competitors. The benefits include time-saving, forexample in avoiding interruptions (the sender can send anytime to the receiver), eliminatingdelays (waiting for the mail), and reducing unproductive face-to-face meetings (time-consumingcoordination effort and absence of some members may delay decisions). E-mail also enablesinternational communication which would otherwise not be possible due to personal reasons ortravel restrictions.

The quality of communication is also improved - permanent searchable records are created,mistaken communications are reduced, and the sender is forced to express his message withbrevity, clarity and precision. Users are also more likely to be less inhibited when expressingtheir opinions.

However, instances of e-mail abuse also abound. "Toxic information" and "junk mail" in the formof pornography, advertisements and even enemy propaganda has brought down somecommercial e-mail systems. Some e-mail systems provide a subscription feature to allow thereceiver to specify only the categories of mail he wishes to receive.

Sproull and Kiesler [SPROULL91] argue that the real potential of network communication suchas e-mail has less to do with faster communication and greater efficiency than with "influencingthe overall work environment and the capabilities of the employees." Networks can be used tofoster new kinds of task structures and reporting relationships. They can change theconventional work patterns of who talks to whom and who knows what.

Stewart Brand in his book, The Media Lab: Inventing the Future at MIT [BRAND87] , notes

The most surprising and consistent quality in e-mail communities is the human warmththey develop...Wondrous subcultures flourish in them as people seek out other peoplewho share their kink regardless of creed, color, address, or time of day...It's the kind ofrevolution the Media Lab is fostering - people taking over the medium in convivialfashion...

Finally, as Ray Stata [STATA92] points out:

There are many impediments to organizational learning, but the most basic iscommunications...Learning to work effectively in teams and groups, which is at the heartof organizational learning, is also shaped by our ability to communicate. Only in recentyears have I begun to fully understand how the words which come out of my mouth andmy pen affect organizational performance both for better and for worse.

As a communication medium, e-mail would also fit in with Ray Stata's comments. E-mail hasgarnered wide acceptance in linking people and organizations in an ever denser web ofrelationships particularly in the corporate and academic worlds. Can exploiting thisconnectedness become a basis for organizational learning in networked corporations?

4.2 Examples of E-Mail Interactions

Sproull and Kiesler also found that "people use technology in surprising ways, and the effectsoften show up that contradict both theoretical predictions and managerial expectations."

An example from Hewlett-Packard illustrates how corporate culture may be subverted. It hasbeen reported that the firm's MBWA (Management by Wandering Around) for maintainingentrepreneurial informality has been partially usurped by MBSA ("managing by screeningaround"). Has HP "learned" more from e-mail interactions than it has lost in reduced MBWA? Isshared vision better built through electronic links than through actual interactions? Anotherinternal study of HP's e-mail systems concluded that the "potential for real change caused by amedium which allows widely separated people to aggregate their needs is, in fact, quitefrightening. Some managers correctly foresee that such a system can be upsetting to the currentestablished order, and do not participate in it as a result" [FELDMAN86].

IBM has long experimented with, and benefited from, computer conferencing and e-mail.However, managing this resource initially raised a host of problems and yielded some valuable

insights into organizational behavior. On the one hand, an expertise network could be rapidlycreated to address some technical problems. On the other hand, the medium was used to bypassthe hierarchy and other established controls. While peer-to-peer communication for informal,and sometimes inappropriate, purposes was widespread, there was initially little feedback tomanagers. With low management involvement, the true value of e-mail was not fullyappreciated. Fortunately, the management yielded to good commonsense by allowing a self-governing user committee to be formed to oversee the appropriate use of the medium. Theexperiment was eventually considered a success. The conferencing system was effective inserving five general categories of uses : (1) appeals for assistance; (2) tips and techniques; (3)software failure reports and enhancements suggestions; (4) product information and reviews; and(5) reviews of internal programs. [IBM881

Tom Peters [PETERS92, p431] also notes the extensive use of e-mail among CEOs. Richard Pogueof Jones Day Reavis & Pogue, a global law firm, uses the medium to keep communicationpersonal, collegiality high, and far-flung offices part of the team. He also gives faster replies to e-mail messages, as opposed to office memos. William Esrey of United Telecom touts e-mail foropening "a dialogue" with employees at all levels, from all over the firm. Esrey claims thatemployees feel they are dealing directly with him. Employees speak differently than if theyknow their messages would be screened by staffers. A recent ComputerWorld survey of e-mailusage by CEOS [COMWORLD93] also confirms this trend.

ARPANET, a large-scale network developed by the Advanced Research Projects Agency of theUS Department of Defense, has an e-mail facility that allows scientists around the country toexchange ideas spontaneously and casually. Heads of research projects also use e-mail tocoordinate activities with project members and to stay in touch with other research teams andfunding agencies. A network community quickly formed, populated by collaborators who mayrarely meet in real life. Scientific opinions and reviews of work done can beexchanged/broadcast to all interested members and critiques can be fedback in a timely manner.A similar situation could equally well develop in any major corporation with international R&Dfacilities. [KIESLER91]

As Kiesler et al point out:Electronic communication differs from any other communication in space, time, speed,ease-of-use, fun, audience and opportunity for feedback. For example, in one firm wheresomeone posted a new product idea on the network, the proposition was sent in oneminute to 300 colleagues in branches across the country, and, within two days, sufficientreplies were received to launch a new long-distance joint project.

Such is the allure of e-mail.

4.3 Sociology of E-mail

E-mail use has been extensively researched but anticipating the effect on communication has beenno easy task. The discussion in the sections that follow draw upon some survey findings madeby M. Turoff et al [TUROFF84] as well as Sproull and Kiesler's studies on how small groups makedecisions using computer conferences, e-mail and face-to-face discussions [SPROULL91 &KIESLER86].

In an electronic exchange, social Cues that influence group dynamics are absent. Electronicmessages lack information regarding job titles, social importance, hierarchical position,appearance and so on. Formal and casual exchanges also look essentially the same.

As a consequence, users are less inhibited in their expressions, leading to a freer flow of ideas.The relative anonymity and loss of status cues lead to a more democratic exchange of views and

even willingness to disclose private information. Everyone appears to have a more equal say.Decisions by majority vote are made with fewer personal constraints.

Teams with different levels of expertise can integrate better without being overwhelmed by theorganizational "pecking order" as social context is weak or absent. The availability of pastrecords of interactions in computer files also enables new members to join in easily and existingmembers to opt out without loss of information.

In providing a substitute for the telephone, e-mail enables users to be more productive, freeingthem from unscheduled interruptions to their thoughts and activities. "Cheap hours" or hourswith low opportunity costs can be used to answer low-priority messages that arrive during peakperiods.

By allowing an individual to communicate informally with distant colleagues, e-mail can helpimprove morale or reduce anxiety/ stress at times of organizational uncertainty.

Users can belong to several social networks concurrently. They are thus likely to be exposed to awider range of intellectual as well as social interactions. Electronic communities are built arounda topic, independent of the work the members are actually doing. Members are likely to bechosen with more regard for their specific expertise and relevance to the decision-making processthan for their location, organizational unit, or place in the hierarchy.

One research study linked the preference for using a particular source of information with actualand psychological costs. These psychological costs relate to the need to reciprocate and to avoidbeing perceived as incompetent by their peers. In a sufficiently large e-mail community, thesecosts are low as there is likely to be a constant recycling of favors with other users that one maynever meet in real life. [ALLEN77, ACKERMAN92]

On the other hand, bursts of anger ("flaming") and disagreements are more likely to occur than inface-to-face meetings. Polarization of views and extreme positions can easily develop in theabsence of non-verbal cues. Many e-mail users supplement their messages with typographicconventions to indicate their moods and other social context, e.g. sad [:-(], joyful [:-)] and so on.

Decision-making can be slowed down in a situation where there is no apparent leader ormoderator and discussions drag on interminably, as may happen in a computer conference. Onthe other hand, it has also been noted that electronic groups tend to be more risk-affine incomputer-mediated discussions than they might have been in face-to-face encounters.

There is also an unwritten norm in many organizations that only hardcopied announcements areofficial, thus undermining e-mail's role.

4.4 Impact on Organizational Learning

For the purpose of this thesis, we shall analyze e-mail as a store-and-forward messaging andcomputer conferencing tool.

We shall confine the discussion of organizational learning to the Divergence and Executionquadrants of our Problem-oriented cyclic learning model. We are not concerned with learningthe functionality of e-mail - which is straightforward enough - but its potential as acommunication medium. Convergence and Assimilation activities are irrelevant to the vastmajority of e-mail users as planning for and implementing an e-mail system is normally the workof professional information technologists.

We shall focus mainly on usage issues. E-mail usage includes exploratory discussions, fact-finding, opinion gathering, and peer-to-peer as well as peer-to-superior interactions.

4.4.1 E-mail and Cyclic Learning

In the Divergence quadrant, e-mail can be used to explore new work opportunities (e.g.telecommuting) and group relationships. New patterns of work and communication aredeveloped and tested. Usage patterns that are sustainable continue to be used while those thatare not are rapidly abandoned. For example, since e-mail allows the user to answer messages athis own time, he is free to redesign his work schedule creatively to take care of messages in orderof priority. Electronic communities that share common goals are likely to survive longer thanthose that merely exchange information on an occasional basis.In the Execution quadrant, e-mail can be used to overcome geographical and organizationalbarriers to knowledge-sharing. For example, e-mail is used extensively to exchange researchideas in the ARPANET and for communication between office workers around the world.Opinions can be surveyed rapidly and decisions can be reached in a timely manner.

In using e-mail, the learning cycle is not complete. The improvement in communication does notdirectly translate into improvement in organizational behavior. Besides, there is also a cost interms of possible reduction in face-to-face contact which allows leadership to be exercised andaction to result.

Concrete

Abstract

Active Reflective

Figure 4.1: Experiential learning in e-mail usage

4.4.2 E-mail and Learning Loops

E-mail provide a single loop solution to inter and intra-organizational communication byovercoming hierarchical, geographical and functional barriers, albeit at the cost of face-to-facecontact. Kiesler [KIESLER86] points out three important social effects : (1) adding new

information in the organization; (2) creation of new groups; and (3) new forms of socialinteractions.

Suppose a group of geographically dispersed executives forms an electronic community tobrainstorm on some persistent problems. E-mail enables this to take place as it allows differentindividuals to work asynchronously at a time convenient to him. Learning is embodied in thecreative utilization of the medium to improve communication and work productivity.

The challenge is to close the gap between the new knowledge and improvement in performance.In hierarchical organizations, even if shared understanding on a solution to a problem isachieved, subsequent corrective actions may be impeded by the bureaucracy. The discussionmay then either continue with added frustration or simply fizzle out.

Double loop learning takes place if conditions exist for, or require, critical analysis oforganizational norms so that renewal of insights can be effected. For instance, e-mail usage mayresult in a radical change in work patterns, e.g., widespread usage leads to bypassing establishedcontrols and hierarchical relationships. Another possible outcome is the expansion intelecommuting made possible by the medium. In both cases, management will have to makesignificant adjustments to their methods of appraising employees, decision making andcoordination. Then double loop learning would be needed. Otherwise, there is little scope fordouble loop learning through routine e-mail usage.

4.4.3 E-mail and the Learning Disciplines

An individual linked to an electronic network is immediately able to communicate with fellow-workers and outsiders on a wide range of topics of concern to him. Learning constraints such asinterpersonal and cultural limitations are substantially reduced in the absence of social cues.Knowledge and expertise can be traded with another person or group with minimal costs.Besides enabling connectedness, uninhibited dialogue with others enable personal visions to beformed and developed. In the setting of a networked organization, these clusters of personalvisions gradually coalesce to form a shared vision within the group, division and corporation.The leader, e.g. the CEO, can use this medium to maintain on-going, good quality communicationwith his subordinates at several levels, further fostering a sense of connectedness.

CEOs can also utilize e-mail to foster personal visions which ultimately become shared visions.Important strategic decisions can be shared with everyone in the organization. For example,when NeXT Computer's top management was trying to decide whether to sell a large stake toCanon, Inc., e-mail was used to obtain feedback from every employee.

The democratic environment of electronic networks also enable team learning to take place. Freeform dialogue leads to creative ideas which may be pursued with a different subnetwork built upfrom shared interests and expertise. Complex problems can be tackled concurrently by a globalnetwork of user expertise. Users have the opportunity to reflect on an issue at their own rate andwith the aid of their computer files

Mental models also may be refined or reshaped. Through interactions with to several different e-mail communities, decrepit worldviews may be modified and replaced by a new set of realities.While the medium can be used to advocate a viewpoint, it also has the advantage of fast feedbackfrom others. Inquiry can take place in an open, partially anonymous setting.

Through e-mail communication, one sees new links and interactions because of the exposure togreater information, discussion and analysis. There is intellectual synergy that encourages newmodes of thought. But this falls short of systems thinking which depends on perceivinginterconnectedness.

4.5 Impediments to Learning

E-mail usage is not commonly perceived as learning, and users will thus not associate their e-mailinteractions as learning. The fragmentary nature of these interactions also preclude a sustainedbuild-up of experiences and knowledge that is likely to have lasting impact on the users' mentalmodels. It has also been found that e-mail interactions encourages rapid exchange of informationas opposed to reflective or critical analysis of issues.

That some learning takes place at all during electronic exchanges is undeniable. What isdebatable is the level and usefulness of this form of learning. It is virtually certain that someindividual learning can be achieved through e-mail communications. But the process throughwhich individual learning advances organization learning is a research topic still in its infancy[KIM92]. From the organizational standpoint, individual learning is irrelevant if not translatedinto organizational improvement.

Electronic connectedness, moreover, has to be complemented by actual face-to-face contact forsustaining the individual's other organizational needs (e.g. recognition, self-actualization).

Another caveat is that the individual might get locked in to his special interest groups anddevelop even more unbalanced worldviews on the basis of support from his e-mail community.

One might also question if e-mail interactions provide a sufficient platform for effectivelypreparing management's mental models for future use, particularly in an environment whereeveryone's e-mail interactions are unique and unrepeatable. The result of such uniqueexperiences may hinder shared understanding.

Systems thinking requires the ability to see through the maze of complexities andinterconnections to what is really essential. Peter Senge [PETERMK92] suggests that learning(systems thinking) may speed up if there is a critical mass of senior executives willing tochallenge top-level strategic assumptions, while groups at local operating levels simultaneouslychallenge operational practices and processes. While computer conferencing provides such aforum, it is difficult to coordinate such an enterprise-wide dialogue. Moreover, the lack of socialcontext and cues may inhibit effective communication, as the receiver may not be able to attachthe right significance to a point raised by the sender.

The inability to support systems thinking means that learning will be at best fragmented. E-mailusers are not likely to feel interconnected with the rest of the organization's processes. Solutionsto problems are likely to be localized in scope.

4.6 Discussion

For a small enterprise, e-mail is a reasonably viable medium for all employees to communicatewith their top management and other peers for knowledge sharing. This expertise network maybe perceived as an informal learning laboratory that facilitates the sharing of knowledge andvisions. However, e-mail users are not likely to perceive their usage patterns as learning sessions.

There is no question that computer-mediated communication will facilitate the transformation oforganizations from a hierarchical to a networked structure. However, e-mail is not designed withthe objective of providing support for organizational learning. Hence its limitations for thispurpose are only to be expected. Even as a communication medium, e-mail cannot fulfill all thepsycho-social expectations of face-to-face group interactions. Learning is more than just acombination of improved communications and knowledge sharing.

Hence, although e-mail does support some of the learning disciplines, its potential as a learningtool is limited. This conclusion flows from the argument that (1) learning is an integrated processinvolving thinking and action; (2) learning is more effective if it is intentional and conscious; and(3) dialogue requires opportunities for reflection. E-mail usage does not necessary tie with action.There is also little intentional learning in e-mail usage. Finally, e-mail is not designed to facilitatedialogue. Computer conferencing provides an environment for group decision making but itseffectiveness is not always convincing.

Furthermore, as noted, e-mail interactions do not facilitate systems thinking. In theabsence of major shifts in work patterns, there is also little room for double loop learning.

Chapter 5: Implementing Groupware

5.1 Overview of Groupware and Collaborative Computing

The invasion of the personal computer into the workplace has raised a chorus of questionsregarding what need to be done to realize the potential of increasing worker productivity andits measurement. In the context of the networked organization, some authors andconsultants are advocating the creation of team-based structures or workgroups through thecombination of collaborative computing and re-engineering of business processes[TAPSCOT93 & BULLEN88].

Workgroups may be defined as collaborating groups of information workers. A workgroupmay also consist of the entire firm where everyone works together to handle the full range oforganizational tasks. Modern electronic communication technologies have eroded theboundaries between firms and changed the nature of their coordination across geographicallocations. Hence, it is not straightforward to identify a formal organizational unit as aworkgroup. People co-located in an office do not necessarily form a workgroup and peoplegeographically separated may form a team to achieve a common work result.

Workgroup or collaborative computing is a catchphrase for a new body of hardware andsoftware that helps people to work better together. A collaborative system creates anenvironment in which people can share information without the constraints of time andspace. Network groupware applications now link workgroups across a room or across theglobe [BYTE93].

The term "groupware" is being popularized for software that is intended for use incollaborative computing. Robert Johansen [JOHANSEN88] gives perhaps the broadestdefinition :

Groupware is a generic term for specialized computer aids that are designed for theuse of collaborative workgroups. Typically, these groups are small, project-orientedteams that have important tasks and tight deadlines...Sometimes, groupware is usedby permanent groups or departments...Group interactions may be formal orinformal, spontaneous or planned, structured or unstructured.

Typically, the goals and tasks of workgroups which may be supported by groupware are togenerate ideas, develop action plans, make decisions, negotiate solutions and refine data, textand graphics.

Recent commercial products include integrated communications environments like Notesfrom Lotus Development and Windows for Workgroups from Microsoft. Other more

specialized but innovative products include Coordinator II, a conversational structuringgroupware from Da Vinci Systems and GroupSystems V, a group decision support softwarefrom Ventana.

Lotus Notes, regarded by some experts as a definitive groupware, is a messaging system builton top of a document database. It is a group communication environment that lets the useraccess and create shared information. Documents stored in databases, which may be locatedat remote sites, can be viewed and edited by several users. Databases and documents can becategorized and searched. As a multi-platform, software architecture on which groupware aswell as traditional applications can be developed, Notes is used for word processing, textretrieval, imaging, bulletin board systems, database links, document management, formsdesign and routing, and workflow.

E-mail is also taking on a smarter, broader role as an enabling technology for a new wave ofgroupware that builds on the messaging infrastructure. The extended role of e-mail nowencompasses communication between people and processes (known as virtual users) andeven among processes themselves. These newer applications include multimedia mail, faxrouting, database access and document sharing. However, the most promising e-mailapplication domain is workflow automation which encompasses information routing, taskautomation, and decision support.

There are presently a large number of collaborative systems with a large variety offunctionality. These systems differ in three fundamental aspects : common task, sharedenvironment and time/space. Common tasking measures the extent to which the membersof a workgroup can work on the same task. For example, a software design system thatallows programmers to interact frequently in real time on a module of code would rank highin the common-task spectrum. A good shared-environment keeps the user informed on aproject's status, who the coparticipants are, and what environmental setting is supported[BYTE93].

Time/space collaborative systems (see Figure 5.1 below) use the dimensions of time andspace to create four domains of collaborative interactions [BULLEN88]:1) face-to-face (same time, same place);2) distributed synchronous (same time, different place);3) asynchronous (different time, same place);4) distributed asynchronous (different time, different place).

The appeal of groupware is that, more than just making existing systems faster or moreaccurate, it provides tools to implement entirely new processes and optimize them to meetorganizational goals. Groupware enables re-engineering of current information systems andmakes workflow automation feasible. Disparate systems may be linked together so as toenhance workgroup dynamics, melding systems and people into effective working units.

Face-to-face Asynchronous

Meeting Facilitation Presentation softwareGroup Decision Support Project Management

System Work Team SupportDistributed synchronous Distributed asynchronous

Conference Calls Calendaring and schedulingVideo Conferencing Computer Conference

Screen Sharing Collaborative AuthoringSpontaneous Meetings Conversational Structuring

Figure 5.1: Types of applications under each time/space domain.

For any workgroup to succeed, communication and collaboration are required. Technicaland behavioral challenges need to be simultaneously confronted. It falls upon groupware tomediate human communications and provide transparent coordination. Hence, groupwaredevelopers and implementers have to think about organizational acceptance issues.Technical quality or innovativeness do not necessarily add up to successful diffusion in anorganization.

5.2 Examples of Groupware applications

Group decision support systems (GDSS) are used to facilitate face-to-face meetings. Theyprovide tools for decision structuring, idea generation, voting and ranking. Besidesproviding anonymity, which improves participation, there is more likelihood of reachingconsensus and commitment to some action plans developed during a GDSS meeting. Apossible pitfall is that formal procedures for decision making are not well accepted by "realbusiness people". [JOHANSEN88]

A session may go like this : a question is phrased to the workgroup's satisfaction. Next, theGDSS asks them for anonymous judgments, certainty and self-rated expertise levels. Whenall the members have input their judgments, the GDSS aggregates their opinions and feedsback a first-round set of judgments from the group. These rounds are repeated until adecision is reached. The GDSS does not make the decision, but provides an effective andefficient group decision making process.

Workgroup schedulers must work with many schedules at once so that it can find the earliestpossible date and avoid conflicts in scheduling meetings for everybody in the group. Apersonal scheduler may provide for scheduling appointments and recurring events, handleautomatic conflict checking, and produce reminders. A workgroup scheduler has to providefor shared updates to personal schedules, user and administrator security, arrange forconflict-free scheduling, sends reminders to participants, tracks meeting attendance,schedules rooms and equipment, and so on. In short, it provides a tool for the workgroup asa whole as well as a personal scheduling tool for individuals in the workgroup.

Work flow is considered a subset of groupware that is fast gaining popularity as a means tore-engineer business processes. Workflow systems use the e-mail transport to route electronicforms, thus saving paper, boosting efficiency and adding intelligence to the data distribution.Intelligent routing is based on a combination of rules, events and time. However, it isimportant to realize that workflow automation systems are designed primarily to deal withthe routine and the predictable. Successful workflow applications employ a variety of tools,including e-mail, scheduling, a shared database, and collaboration tools, all accessible fromthe same platform.

An example of work flow implementation that requires cross-disciplinary collaboration isClorox's Manufacturing Standards documents distribution system [SEYBOLD92]. Theproblem was that there were some 8000 manufacturing standards documents (technicalcommunication from R&D to manufacturing on process, tests, materials, etc.) that have to beaccessed by multiple plants in several different countries. On average, there are 400 changesa month, and 10000 pages are distributed. These manuals and changes also had to be typedup and printed following a correct format.

The cross-functional team identified the goals as: (1) reduction in cycle time from author touser; (2) improvement in standards concurrence; (3) collaboration support, e.g., feedbackfrom plants to authors; (4) ubiquitous document access ; and (5) extensibility to differentclasses of documents. The team recommended the use of Lotus Notes for its document-handling and information-sharing capabilities as well as its ability to facilitate fast two-waydocument access. The number of retrieval and feedback steps were reduced by eliminatingneedless loops in the document flow. Related drawings and documents are now linkedtogether for fast and easy retrieval. The implementation team also reported side benefitssuch as the creation of lateral communication channels and cross-organizational pull. Newincentive and appraisal systems were introduced to encourage the use of the system.

5.3 Organizational and cultural issues

Studies by Wanda Orlikowski, Christine Bullen and John Bennet, Patricia Seybold and others[WANDA92, BULLEN90, SEYBOLD921 have shed light on a number of organizational andcultural issues that impact the implementation and acceptance of groupware.

Groupware is found to be more successful at lower levels of an organization. These lower-echelon employees are used to team projects and are willing to remain as anonymous sourcesfor comments, ideas and suggestions. Technologists who are not subject to competitiveculture, or individual-focused reward systems, appear to be comfortable in using thetechnology to solve technical and work problems.

Groupware also requires more commitment. A workgroup scheduling system requires 100percent from all the members in order to work. Every member must faithfully updatehis/her personal schedule and make his calendar shareable to some extent. Individualpreferences may have to be given up for the sake of the group. Some members may feel thatothers are benefiting more from their work or sacrifices.

Some groupware functions actually add more effort on the part of the users to learn to dothings in a different way, independent of the technology. People may not be willing to adapttheir behavior and invest the personal resources necessary to take advantage of some of thefunctionality.

Examining people's attitudes appears to be an important step in understanding barriers toproductive work. When people saw the immediate value to themselves of using a functionthey quickly adapt to its use. However, when it was in the interest of the team, department,or organization, individual incentive was usually lacking.

Most people use what functionality they had been taught in formal training sessions withoutputting in the effort to use the tool creatively to carry out their work. Only when concepts areunderstood do people begin to apply creativity to using the functionality.

Managerial support cannot guarantee successful implementation by itself. Managementmust carry out specific tasks to bring forth productive team characteristics. Management

must also set the right expectations in order to derive ultimate acceptance. For example,when introducing the groupware, they should emphasize the critical impacts on theemployees' work performance.

When management held high expectations for productivity enhancement throughgroupware, yet gave no attention to examining the work process, people reported they feltunder pressure to perform better by learning a new tool without a clear understanding ofhow the tool could make a difference.

Other factors that influence groupware acceptance relate to culture, e.g. if the organizationpromotes competitive individualism, it is likely to conflict with the premise of groupware -sharing, collaboration and participation.

For collaborative systems to be truly successful, business processes must be changed. Pre-groupware reward systems will be a major hindrance to the acceptance of the tool.

Security may be an issue in shared data. New policies need to be enforced on data quality,confidentiality and access control.

Much of the success of teams, and therefore of groupware, depends on communicationamong team members. Organizational success factors include: 1) clear elevating goal(s); (2)results driven structure; (3) competent team members; (4) unified commitment; (5)collaborative climate; (6) standards of excellence; (7) external support and recognition; and (8)principled leadership. [BULLEN90, LARSON89]

5.4 Impact on Organizational Learning

For most groupware applications such as workgroup scheduling, collaborative authoring,etc., the extent of learning that results from usage would not be radically different than thatof e-mail interactions. Hence, we shall focus the rest of the discussion mainly on workflowimplementation and usage.

5.4.1 Groupware and Cyclic Learning

The context of our discussion will focus mainly on the Assimilation, Convergence andExecution quadrants of the Problem-oriented cyclic learning model. Although we shall focuson these two quadrants, a workflow automation project does actually complete the learningcycle of feasibility study, planning, implementation and usage. By completing the learningcycle, the potential for learning is significant, particularly if there are opportunities for futurere-cycles in which new insights may be discovered.

In the Convergence quadrant, the activities center on implementation strategies for aworkflow system that parallels the current or re-engineered process. Other issues includeworkers' training, and changes in work practices and reward systems. As a relativelysignificant amount of the creative thinking (knowledge creation) is likely to emerge in thisquadrant, we expect significant learning to occur here as well.

The Execution quadrant is relatively static. The users learn to adapt to the new process.Once implemented, the groupware application is likely to become routine except for someoccasional problems caused by external or unpredictable circumstances. For non-workflowapplications like GDSS, workgroup scheduling, etc., some learning may take place in thisquadrant.

The Assimilation quadrant is included because planning is required for making the transitionfrom a manual to an automated system. Some planning is also required for the workpractices and reward system to be revised to take cognizance of the change in the businessprocesses in order for new problems to be resolved satisfactorily. However, these changesare more likely to be reactive and implemented mostly at the operational level asmanagement adjusts to the new workflow system.

FeasibilityConcrete Usage Study

Divergence

Execution- adapting to new

system Assimilation- reacting to

unexpectedproblems Convergence

- integrating businessand technical knowledge

Implement - designing new systemAbstract ation Planning

Active ReflectiveFigure 5.1: Experiential learning in groupware implementation

5.4.2 Groupware and the Learning Loops

Single loop learning may be defined as improvement (more of the same, but better).Implementing a workflow system that parallels the existing one therefore is a single loopsolution. Changes are only incremental. There may be minor adjustments to work processes.

On the other hand, reengineering the business process in conjunction with workflowimplementation would be a double loop learning process as it results in renewal of insights,understanding and knowledge about the process and application area. Somewhat moreradical changes to the organizational culture are required to accomodate the new workprocess.

With a flexible architecture like Lotus Notes in place, it is possible for the workflow processto be reengineered repeatedly with minimal reconfiguration of the information resources. Itis thus foreseeable for the reengineered process to continue evolving in incremental steps, i.e.,going through a series of single learning loops or improvements, before a major overhaul(double loop solution) is required. Re-designing new reward systems and work processesalso result in knowledge creation which leads to both single and double loop learning.Application of systems thinking will help cut down the number of "deadend" single loopsolutions.

5.4.3 Groupware and the Learning Disciplines

During the implementation phase, there is ample opportunity for increasing competence andresolving creative tensions in working with a cross-functional team set up to tackle areengineering workflow automation project. The mental models of the team membersundergo changes as they observe analogies within their own departments. Team learningtakes place through the dialogue and coordination inherent in a cross-functional project.Team learning, mental models and shared vision building are all realized and supported at afar more intense level than in the case of e-mail or groupware usage. Similarly, there is morescope for personal mastery - personal visions for the new work flow process can be sharedwith team members. Creative tension is resolved through experimenting with, and makingimprovements to, the work process. When the workflow system is in use, the changed natureof work may need a new reward system and hence more learning is required.

Cross-disciplinary teams contribute insights from their experience and knowledge of theirown areas. Implementers act as peer reviewers, providing an objective understanding of thesystemic interplays. Everyone goes away with a better picture of the organization and thework processes involved.

Systemic thinking emphasizes that "structure influences behavior" - the new workflow willchange work patterns and hence influences employees' behavior. Systems thinking alsorecognizes that there is often resistance to change - "policy resistance". Finally, a fundamentalgoal of systems thinking is to identify the leverage points in a complex problem so that, bysolving those critical subproblems, the overall problem is resolved. A systemic solution maynot necessarily be the result of the implementation process.

In implementing a workflow automation solution, the implementation team must understandthe shortcomings of the present solution and try to rectify by putting in place an improvedwork process with the help of new IT tools. Training and employees' attitudes also need tobe considered in order to introduce the new workflow successfully. Users should not beexpected merely to follow the new routines unquestioningly. They should be trained andencouraged to provide creative innovations, if necessary, that improves their work.Management must set their expectations clearly and not try to push the groupwaretechnology for its own sake.

For example, when Clorox decided to implement the document flow system, a cross-functional team was assembled to evaluated various approaches and tools both for thepresent as well as future requirements. The "leverage" points were identified as (1)duplicated efforts; and (2) long flow loops resulting in unnecessary delays; and (3) multipleaccesses required to retrieve related documents. The solution was further constrained by theexisting infrastructure consisting of mainframes, networks and personal computers. Manydepartments (Manufacturing, R&D, Quality Assurance, international divisions) would beimpacted by the new workflow. Systems thinking is needed to evaluate alternative strategiesin such a complex problem.

5.5 Impediments to Learning

Groupware challenges the organizational culture with a new means of communication. Itintroduces a new on-line culture. Groupware supports shared effort, cooperation, andcollaboration. If these values are antithesis to the organization's culture (competitive &individualistic, rigid hierarchy etc.), the technology will not be well accepted. This istherefore a cultural limitation which impedes learning.

People are hindered in the use of the tools because of problems associated with the degree ofintegration. People who have worked only in PC environments may face cognitive limitationin grasping the cooperative nature of the applications. In such cases, learning groupwarecollectively may foster joint understanding and shared expectations.

Interpersonal and cultural limitations surface in the form of competitive individualism whichreinforces individual effort and ability but does not support cooperation or sharing ofexpertise. Frequently, it is upper management which faces disincentives for sharing ideasand data - personal information is a powerbase. Competitive culture also forces managers towork alone.

In some workflow systems, there is a tendency for social distortion whereby workers whoroutinely meet work targets are rewarded while others who spend time on difficult customerproblems are penalized. Electronic messaging would provide the forum for airing suchgrievances. More fundamentally, the reward system must be modified to correct thisdistortion. The new reward system should be team-oriented as opposed to individualoriented. Reengineered systems must value collaboration, worker involvement and moreparticipative structures. Otherwise, the advantages of groupware would be lost and teamlearning impeded.

5.6 Discussion

Implementing groupware provides opportunities for experiential learning in the form ofcross-disciplinary sharing of knowledge and experience in designing a new work process.Typically, in workflow automation problems, double loop solutions are called for as the workprocess has to be reengineered. Creative and critical thinking is required to identify the keyissues and design a new work process. The reward system also needs to be revised tosupport team goals. Workflow automation solutions are systemic in nature; hence systemsthinking should be applied.

Learning impediments typically relate to the existing culture which rewards individual asopposed to team efforts. Resistance among the would be users may surface due toinadequate appreciation of the technology's potential and misunderstanding managementintentions. To overcome these impediments, management should come to a shared visionwith the employees involved. The old mental models of computer technonlogy and modes ofwork need to be revised. It is also worthwhile for management to initiate a dialogue (teamlearning) with the employees with a view to revamping cultural and organizationalconstraints like an archaic reward system.

Therefore, groupware implementation holds significant promise for enabling organizationallearning by providing a platform for cross-functional collaboration. Team learning, mentalmodels, and shared visions are well utilized. Personal mastery is possible both in the form ofcompetence enhancement and providing creative inputs into the project. All these disciplinesare translated into action when the new work process is implemented. Hence, the learningexperience is far more intense than in the case of e-mail usage.

Chapter 6: IT STRATEGIC PLANNING

6.1 Information Technology and Paradigm Shifts

Don Tapscott and Art Caston, in their book, Paradigm Shift -The new promise of informationtechnology - identify eight technology shifts that distinguish Era I from Era II computing

[TAPSCOTT92]. These eight critical technologies may be broad grouped under networkcomputing, open systems, and software revolution.

Network computing is made possible through the superior price/performance of themicroprocessor over the traditional mainframe. In moving processing power out to wherethe action is in the enterprise, the microprocessor-based personal computers or workstationsempower the individual worker. In order to share data and other IT resources, theseworkstations are linked together to form networks. Hence Era II computing is first andforemost characterized by the shift from mainframe host-based to network-based computing.

To achieve effective network computing, the business must develop an enterprisearchitecture. An enterprise architecture is the underlying framework which defines anddescribes the technology platform required by a business to attain its objectives and achieve abusiness vision. It is the structure given to information, applications, organizational, andtechnological means and to the groupings of components, their inter-relationships, and theprinciples and guidelines governing their design and evolution over time. [TAPSCOTT92, p162]

Network computing inevitably leads to the adoption of industry standards, i.e. thedevelopment of open systems. The X/Open consortium, a worldwide organization ofinformation systems suppliers and users, defines open systems as "systems and softwareenvironments based on standards which are vendor independent and commonly available."The two main objectives of open systems are : (1) portability, which enables softwareapplications and information to be easily moved to computers supplied by different vendors;and (2) interoperability, which allows sharing of resources, data and software applicationsbetween different computers which may or may not be linked in a network.

Open systems have been found to reduce IT costs in the areas of hardware, software,information management, and the human costs of managing change. Secondly, open systemsalso provide value-added benefits such as reduced risk of vendor dependence, architecturalflexibility, better software integration, easier migration to new innovative technologies, andwider choice of packaged software. In Era II computing, open systems lay the foundation forvendor-neutral software standards, the integration of data, text, voice and image intomultimedia technology, and the rise in multi-vendor partnerships.

The shift to graphical user interface, the acceptance of the object-oriented paradigm thatimproves module reusability, improved computer-aided software engineering (CASE) tools,and the growth in integrated software applications make up the software revolution.

In Era II, network computing replaces the old host-based paradigm built around large,centralized data centers. A network computing environment provides the means for users toaccess a wide range of information, applications and computing resources without the needto know where and how they are interconnected. Network computing also involvescooperative processing in which application components spread across multiple platforms inthe network are able to work as an ensemble. Network computing is embodied in the client-server approach in which the client (application program) requests information andprocessing services from one or more servers (information sources) in the network. Theclient and server are distinct hardware components with their own processing capability. AnEra II architectural environment treats hardware and software as equivalent.

6.2 Overview of IT Strategic Planning

Information technology and information systems (IS) planning methodologies have beenevolving in tandem with business strategy planning. These methodologies attempt to

identify strategic business opportunities based on IT/IS and to integrate the two types ofplanning. The IT infrastructure is viewed as a strategic resource. Hence, the successfulexploitation of IT requires an alignment of strategic context and the IT infrastructure.

Over the years, the IT infrastructure vision has evolved from being independent on, to beingreactive to, and finally, interdependent of the strategic context [VENKA91]. In theindependent vision, the IT infrastructure plays a support role and thus has little impact onshaping or implementing business strategies. In the reactive vision, the business strategydirects the shape of the IT architecture, but not vice versa. Finally, in the bi-directional,interdependent vision modifications in the IT infrastructure signal implications for possiblechanges and improvements in business strategy, and vice versa.

Basically, there are two families of approaches to IT/IS planning and infrastructuredevelopment. The first approach attempt to use a formal planning procedure sometimesknown as enterprise architecture planning (EAP). EAP is the process of definingarchitectures for the use of information in support of the business and the plan forimplementing those architectures (SPEWAK931. The second approach, recognizing thepitfalls inherent in the first approach, emphasizes on fostering an appropriate climate inwhich strategic applications congruent with the business mission will naturally emerge.While the second approach does not necessarily forego the EAP, its underlying premise isthat IT strategy is driven by continuous re-alignment to users' needs which are shaped byday-to-day operations as well as changes in business strategy.

The first approach, besides consuming time and resources, also tends to be heavilymethodology-oriented. The result is a document known as the IT strategic plan which, morelikely than not, is rapidly obviated by the changing business environment and technologyparadigms. The second approach, on the other hand, attempts to create a shared vision withlead users. Over a period of time, the IT infrastructure and users' goals spiral inwards to theshared vision in incremental steps. The actual planning process is abbreviated, giving scopefor active experimentation with users' participation. To achieve this alignment, the strategiccontext has to include the IT function as an equal partner together with the other functionslike marketing, manufacturing and so on.

In general, however, both approaches have met with partial success. According to a study byDMR and Cognitech [TAPSCOTT92, p190], there is an inverse relationship between formal,separate IT plans and the degree of IT contribution to the business. That contribution couldbe increased by ongoing contact between business and IT professionals and managers, ratherthan through highly structured, formal plans.

Networked organizations have to address global markets which are fast-paced anddynamically changing. The extended nature of the corporation includes suppliers andcustomers who also have a need to access information as much as the internal users. Theseorganizational factors call for a distributed and real-time approach to IT infrastructureplanning and design. At the same time, information industry is also extremely innovative,bringing forth both new advances that supplant current technologies and new paradigms forconceptualizing users' needs. The answer is to create an open distributed environment inwhich new technologies may be readily grafted on and yet flexible enough to accommodatedrastic rightsizing (or downsizing) trends. Finally, competitive pressures imposes the needfor flexible cost structures that is attainable through small adaptations to the IT infrastructureas opposed to a "great leap" as might be suggested by the formal planning approach.

The IT infrastructure is too important to be left to the IS function alone. Lead users have to bedirectly involved through formal planning processes to map the key corporate strategicthrusts onto some tangible IT action plan. As pointed out in Chapter 3, an IT management

architecture comprising of IT and line managers has to be formed to oversee theinfrastructure development. The right information politics, notably federalism, has to befostered to produce a shared vision through negotiation and continuous learning.

6.3 Two Models of IT infrastructure

In most organizations today, host-based legacy applications co-exist with network-basedsystems. In one major financial services company codenamed Finserve, the back officeadministrative functions are host-based whilst the dynamic trading room systems arenetwork-based. The trading room systems are required to process massive transactions inreal time. The needs of the financial services users, both external and internal, evolve at arapid pace, quickly overtaking the capabilities of the existing infrastructure. In the long term,Finserve's chief information architect foresees the two infrastructures melding together.

The administrative information system, conceived in Era I, uses a "data centric" model of ITinfrastructure (see Figure 6.1). The three-tiered framework is structured into the strategic,tactical and operational layers which support a parallel hierarchy of IS managers,professionals and users. Enterprise architecture in the sense of a disciplined organization ofinformation technology implemented to realize a business vision may not really exist here.Rather, this infrastructure is the historical cumulation of various systems implemented overthe years to address the business requirements of the day.

In the data centric model, data is the key asset which has to be modeled, structured,managed, enhanced and updated. The database virtues of non-redundancy, integrity,reliability, and so on are the guiding principles. Data is synchronous, only one version of livedata is possible at any point in time. This well-defined approach usually results in aninfrastructure that is relatively easy to operate and manage. Applications requiring on-linedata access by a large number of users but little processing power may find this model to berelatively cost-effective. On the other hand, this infrastructure is not readily adaptable tochange.

In contrast, the trading room systems, described as "message centric", represent anamorphous, asynchronous, business valued-added model of IT infrastructure. Optimizinginformation access and user empowerment are viewed as the guiding principles. Thearchitecture is implemented as a modular set of loosely coupled subnetworks each of whichserves a specific user application domain. The solution, which is an integration of severalopen platforms, subscribes to the criteria of functionality, performance, availability, cost, easeof operation, and ease of management. These goals are not necessarily congruent and maynot be ultimately realizable concurrently. The non-hierarchical structure fostersinnovativeness and user participation. Line managers work together with IS managers todefine their shared vision. This does not mean that formal planning is altogether eliminated.Planning is less extensive but more participative and responsive to the business environment.

Constructing a message centric infrastructure is not straight-forward. Besides theorganizational issues, the technical challenges are formidable. (Finserve employs about 3000IT professionals to support the two infrastructures.) Since data is distributed or replicatedacross several subnetworks, there may be more than one live version of the same data at anypoint in time. The solution approach is "lazy synchronization" - propagating the updateddata to all the relevant databases in the other subnetworks. The network has to guaranteethat a specific database is able to provide the most up-to-date data when needed. Hence,there are major technical hurdles to design a reliable, fault-tolerant network, otherwise dataintegrity would be compromised.

The chief information architect has to be knowledgeable on technological developmentswhile staying close to the users. He perceives his role as factoring and synthesizing users'needs into a coherent IT framework that can deliver the desired functionality andperformance. In attempting to satisfy the diverse needs of disparate groups of users, anumber of distinct, localized IT networks are first constructed. Over a period of time, eachnetwork evolves at its own pace, maintaining alignment with the application's requirements,and is continuously re-configured to steer towards the shared vision. The architect also playsthe role of a coach and visionary whose responsibilities include conceptualizing, articulating,and re-adapting the users' mental models on the opportunities and threats posed bytechnological advances.

For the network organization, theevolutionary approach to buildingrelationships emerge.

message centric model provides the flexibility andIT infrastructure as new business requirements and

Data Centric ModelHierarchic,Synchronous,Information systems focused

Message Centric ModelAmorphous,

Asynchronous,Business value-added

Figure 6.1: Two IT infrastructure paradigms

6.4 Impact on Organizational Learning

6.4.1 IT Planning and Cyclic Learning

IT infrastructure planning essentially focuses on the feasibility study and planning stages ofthe Problem-oriented learning cycle. However, some of the plans will be pursued through toimplementation and usage stages. Each stage creates new knowledge about the organizationin terms of work processes and business policies, for both internal and external users. Ingoing through the cycle, cross-disciplinary interactions between users and IT professionalsresults in knowledge creation and organizational learning. Although we shall focus on theAssimilation and Convergence activities in this discussion, any learning in the other twostages will have an impact on the next round of planning. Figure 6.2 illustrates this cycliclearning process.

Concrete

intrastructure \

Implement- - exploiting interdependenceAbstract . Planning

ation Planning

Active lop.ReflectiveFigure 6.2: Experiential learning in IT strategic planning

In view of the long range significance of IT planning, learning is expected to be double loop,i.e, double loop planning. Users, who should claim ownership of the results of the planningeffort, and IT professionals worked together to explore new opportunities and to develop ashared vision. Planning now becomes a continuous, action-centered form of learning. Andthere must be follow-up action - the involvement of users in the subsequent implementation.Through this process of thinking and doing, managers learn about current policies andprocesses. This learning is then used in re-designing new and innovative solutions toproblems.

Cyclic learning - which is experiential and action-centered -may be achieved in both the datacentric and message centric infrastructures (i.e. in both Era I and Era II environments).However, in a hierarchical organization which is likely to embrace only the data centricmodel, the presence of a deeply nested organization structure would hinder effectiveplanning by encouraging departmental level planning which are likely to result in parochialviews of the IT infrastructure. This would lead to conflicts in prioritization of IT resources.To avoid this, explicit cross-functional planning should be actively practiced. In thenetworked organization, the flat structure supports the message over the data centric view.There is more scope for new technology to be incorporated, thus intensifying the need formore learning to exploit the new opportunities presented.

6.4.2 IT Planning and the Learning Loops

Single loop solutions provide incremental improvements to organizational behavior bystraightforward automation until such a time when such changes no longer suffice to solvethe problem. Double loop solutions, representing renewal of organizational behavior bychanging the work processes, then has to be sought.

Planning in the data centric model may be conceived as a double loop planning followed by aseries of fine-tuning single loops. When a threshold is reached that requires a major shift inbusiness strategy, a new double loop planning process is started, and the cycle repeats. Ateach solution step, the corresponding learning loop takes place. However, the delaysinherent in implementing a complex, detailed plan are likely to obsolesce the IT strategic plan

long before the next round is due. Figure 6.3 shows how IT planning impacts learning in thedata centric model.

Strategic Thrusts andUser Requirements

Enterprise Architecture(and Revisions)

Double LoopLearning

Single LoopLeaming

Figure 6.3: Learning in Formal IT Strategic Planning

In contrast, planning in the message centric model typically starts with a triple loop solution.The IT function (re)defines its role vis-a-vis the users, forming an IT management architectureas necessary. A shared vision is created from the shared mental models of the participants inthe IT management architecture. Next, one or more single/double loop planning processesmay be started concurrently for the various groups of users. The IT infrastructure iscontinuously re-adapted and re-configured in a series of single loop plans interspersed withoccasional double loop plans. Finserve's chief information architect refers to this process asthe "application life spiral". Figure 6.4 illustrates how the application may spiral into theshared vision.

D

D

Shared

VisionS

D D- 1 D

D

START

S=Single LoopSolution/LeamingD=Double LoopSolution/LeamingT=Triple LoopSolution/Learning

Figure 6.4: Application Life Spiral and Learning

In fact, the "methodology" that Finserve uses for its IT planning is patterned after theStrategic Alignment Model (SAM). [VENKA90] The SAM involves at least the four domains: (1) business strategy; (2) organizational infrastructure and processes; (3) IT strategy; and (4)IT infrastructure and processes. The SAM argues that a strategic planning process mustaddress both strategic fit (internal and external alignment) and functional integration(business and IT alignment). The SAM provides the framework to define four attributes thatcharacterize IT planning effectiveness, namely, consistency, completeness, validity andcomprehensiveness.

Using the SAM, single loop planning is characterized as complete and invalid while doubleloop planning is complete and valid. In other words, double loop planning requiresexplicitly examining the interactions among the four planning domains (i.e. the completenesscriterion) and examining the underlying assumptions (i.e. the validity criterion). Of all thepossible combinations of the above four attributes, double loop planning is considered themost effective. [HENDERSON89] Under conditions of high environmental uncertainty suchas Finserve's trading room systems are experiencing, a high level of comprehensiveness in theplanning process is not necessary.

Hence, Finserve has in fact confirmed the validity of these propositions by the use of the"application life spiral." This approach acknowledges the environmental uncertainty andreduces the need to collect details (high comprehensiveness) during the planning process.The application life spiral shows a series of single loop planning interspersed with twodouble loop planning. This suggests that the trading room systems go through cycles ofstability and uncertainty.

The backroom adminstrative system, on the other hand, is relatively stable. Hence, singleloop planning is often sufficient. However, in time to come, a double loop planning processhas to be initiated to re-align the IT strategic plan in order to "correct" for the increasing lossof validity.

Hence, both data and message centric models provide opportunities for organizationallearning. A major paradigm shift necessitates at least double loop planning while minoradaptations only require single loop learning.

6.4.3 IT Planning and the Learning Disciplines

The information architect must possess both technical competence and interpersonal skills.He has to be able to sense business opportunities presented by new technologies and to findways to reach the shared vision from the current infrastructure. The information architectplays an active role in building shared mental model and vision. Team learning is aprerequisite for a shared vision to be created. This takes place through negotiations anddiscussions with the lead users on projects prioritization, resource allocation and sharing.Business processes and policies may be redefined. The readiness of the users for a specifictechnology also has to be considered.

In designing an IT infrastructure, tradeoffs must be considered, e.g. centralized versusdecentralized processing, control versus user empowerment. These are systemic issuesrequiring systemic solutions. The dynamic nature of the business environment favors adecentralized infrastructure whilst operational conditions and economies may favor acentralized approach. The present and anticipated rate of change of communicationstechnology is another factor. External requirements from customers, suppliers and partnersneed to be integrated. Delays are also virtually inevitable during implementation as new

technological advances may be needed and take time to debug. The information architect hasto bear in mind that the IT infrastructure will ultimately influence organizational behaviorand competitiveness. A less than optimal design will face resistance in its use and requiresubsequent re-engineering at great costs. Systems thinking provides the framework forunderstanding and resolving these interlocking problems.

In general, planning for the data centric model is long-term. Hence systems thinking wouldbe needed to understand the impacts of phased implementation and delays on the strategiccontext. On the other hand, it would also be useful to learn more about how the emerginguser-empowered structures impact policies in the message centric model. How aninformation infrastructure may be leveraged for maximum strategic advantage is a systemsthinking problem.

6.5 Impediments to Learning

The data centric model is the result of a methodology-based planning process. The goals areprecisely stated and the evolutionary steps well-defined. This approach is appealing - thehierarchical model and planning methodology simplifies the otherwise formidableintellectual challenge of envisioning a comprehensive infrastructure. But learning requiresmore than the sterile output of such a formal planning process. There has to be action andparticipation. In formal planning, the planners and the doers are likely to be different groupsof people. The result is that the implementation team lacks appreciation for the goals and theplanning committee neglects implementation issues. In other words, learning is impeded.To overcome this problem, a core IT strategic planning group should follow the plan throughto its implementation. Line managers whose functional areas are impacted should bemembers of this core team. By taking ownership of a project, the user is in a better position tolearn through the entire cycle.

The shift from Era I to Era II (i.e form data centric to message centric model) requires bothunderstanding of the technology as well as the business issues. Attempting to imposetechnological solutions on organizational problems is at best a single loop solution that maybackfire. Similarly, to modify business strategy without taking advantage of theopportunities of the new technologies is likely to be self-defeating. There is significantcognitive limitation in management's ability to grasp the implications of this paradigm shift.Systems thinking should be helpful in overcoming this limitation.

The paradigm shift in technology will work its way down into the corporate culture. Forexample, the organization now has to adopt a flatter structure in response to the increaseduser empowerment afforded by information technology. The organization is also becomingmore "real time", highly responsive to the demands of customers, suppliers, partners andinternal functional units. Management and workers have to learn to adapt to the culturalshift. Hanging on to the old reward and appraisal system would impede change and henceorganizational learning.

Cross-disciplinary and cross-boundary collaboration in planning requires effective teamlearning. Interpersonal limitations may be encountered and have -to be resolved throughdialogue.

6.6 Discussion

IT strategic planning offers wide opportunities for line managers to integrate their visionswith IT managers. The more traditional data centric model of IT infrastructure encourages abroad overview of both business and technological issues before crystalling on an

architecture. The opportunity to incorporate systems thinking and double loop learningexists here. However, the final output is often overtaken by external events, thus frustratingthe learning process.

The message centric model, on the other hand, expects learning to take place continuously.Each incremental planning step is followed closely by implementation, with occasionalcorrections (double loop solutions) to re-orientate towards the shared vision. Learning ismostly experiential as few, if any, roadmaps exist to point the way ahead. Arguably,however, the spiraling steps continuously enforce a team learning environment (i.e. IT-linepartnership) in order to move forward. Ultimately, the shared mental models between theline and IT managers merge into the shared vision.

Summary of Part II

Chart S2.1 below summarizes the level of learning that may be supported in each category ofprocesses and disciplines. The scores are purely judgmental, of course. The assumption is thatideal learning conditions prevail. The final overall score for organizational learning is mainlybased on the scores for systems thinking and double loop learning.

IT Processes Using Implementing PlanningElectronic Groupware InformationMail and (Workflow ArchitectureGroupware Automation)

LearningDisciplines &Processes

Experiential * **Learning

Personal * ***

Mastery

Building **

Mental Models

Building **

Shared Vision

Team Learning ** **** ***/****

Systems **

Thinking

Single Loop *

Double Loop * *** ***

Triple Loop * * ***

Individual ** ***

Learning

Overall *

Organization-al Learning

Score Indicators :* Little or none

SignificantChart S2.1:

Comments

Fairly significantVery Significant

Relative Learning Potential in IT processes

Chart S2.1 shows that, in general, a user who only participates in the Execution stage (i.e whosecontact with the computer is restricted to interactions at a terminal or workstation) is notexpected to learn as much about organizational issues and processes as compared to another userwho takes an active part in the planning, design and implemention (i.e., completing the fulllearning cycle) phases of IT activities. This conclusion flows from the argument that (1) learningis an integrated process involving thinking and action; (2) learning is more effective if it isintentional and conscious; and (3) team learning, which involves dialogue, results in deeperunderstanding. In contrast, group decision support systems and modeling using systemsdynamics tools are executed within a team learning environment, providing ample opportunitiesfor dialogue. In addition, decisions and insights gained during these sessions are givenopportunities to be tested out in real life situations. E-mail interactions fail on these scores.

The Assimilation and Convergence phases provide wide opportunities for systems thinking anddouble loop learning. Hence, by getting involved with planning and implementation activities,the user (functional line managers and their delegates)'learn in terms of mental model adaptationand building shared vision of the business together with the IT function. Team learning takesplace through the dialogue and negotiation processes that inevitably occurs in a federalist,networked environment. Learning is more effective if the users consciously set the goals assuch. More likely, they are indifferent to such goals. Hence, the potential for learning is usuallyunrecognized.

The overall evaluation shows that using e-mail provides limited learning opportunities.Groupware such as group decision support systems and computer conferencing are somewhatbetter at bringing out viewpoints for discussion. New ideas are expected to surface and thegroup as a whole gains a better understanding of the organization's goals and processes.Implementation of workflow automation provides stronger learning opportunities as the abovediscussion explains. Finally, planning allows cross-functional teams to learn collectively.Planning followed closely by implementation (thinking and action) is in fact the best way tolearn if the organization can spare the resources for these purposes.

Implementation and planning activities provide opportunities for team learning, re-adaptingmental models, building shared visions and systems thinking. However, in reality, little learningactually takes place partly due to the constraints of the organizational culture and partly to theindividual cognitive limitations in understanding the complexity inherent in a large scale, multi-disciplinary project. Dynamic organizational environments in which teams split up uponcompletion of a project also make it difficult to accumulate a core of focused knowledge on aspecific organizational activity. Furthermore, the reward system sometimes work in favor ofthose engaged in routine as opposed to innovative thinking.

IT strategic planning in many cases suffers from the inability of IS managers to build a coherentshared vision with users, along with a host of other organizational and technical considerations.Consequently, when the IT strategic plan has to be reviewed, a set of new modifications aremerely imposed on to the existing architecture. Hence, frequently, only single loop learningresults. The organization fails to benefit from the synergy that may otherwise be derived fromthe interdependent vision of IT with the business strategic context.

In order for learning to take place, the learning disabilities in implementation and planningactivities should be analyzed, e.g., using system thinking tools. By being aware of thesedisabilities, the implementer or planner can make learning intentional. The five disciplinesshould be utilized to overcome these disabilities.

Conclusions

I. Review

This thesis has integrated a variety of ideas from several leading authorities on organizationallearning. The cyclic learning model has been proposed and used to highlight learningopportunities at various IT processes (namely, usage, implementation and planning). Learningopportunities can be impeded by learning disabilities. Peter Senge, however, believes that thefive learning disciplines of the learning organization can act as antidotes to these learningdisabilities. [SENGE90, p261. The problems encountered in the IT processes studied are triggersfor different levels of organizational learning which may be categorized as single, double andtriple loop learning or solutions. Each of these learning loops produces results which impactorganizational behavior and leads to organizational change. Single loop learning results inimprovement; double loop learning results in renewal and triple loop learning results indevelopment. Finally, the five disciplines, particularly systems thinking, serves to underline theessence of learning in the IT processes.

In Chart C.1, the learning opportunities and disabilities/impediments in IT processes arepresented. The contents are a condensation of the discussions in Chapters Four, Five and Six.

Chart C.2 shows how the learning disabilities may be countered by the five learning disciplines.The role of systems thinking in IT implementation and planning is critical. Systemic issuessurface repeatedly in any planning activity. In IT strategic planning, syetems thinking is neededto analyse and understand complex work processes involving several divisions or externalpartners; to evaluate ways to integrate IT activities with business processes; and to anticipate theimpact of technological change on the organization so as to be able to create the rightinfrastructure.

It can be seen that learning opportunities exist in all the quadrants. Learning impediments ordisabilities are also present at every stage. A networked organization, which is heavilydependent on the IT infrastructure, will find itself in a favorable position in attempting to create alearning environment. By awakening to the learning opportunities in IT, the employees andmanagement can directly benefit from the vast wealth of experience and knowledge that can begained from cross-functional, cross-boundary activities such as those which are likely to becreated among suppliers, customers and partners within the extended domain of the networkedorganization. Such experience can be diffused throughout the organization through effectivecommunication, team learning and shared visions between employees.

Chart C.3 shows the relationship between IT and the learning loops. The results of learning arealso shown.

It can be seen that even single loop learning can produce good positive results such asincremental improvements to business processes. Double loop learning results in a more radicalchange - a renewal of the goals and orientations. Double loop solutions are usually attemptedonly when single loop solutions fail to make progress at resolving a problem or error condition.For example, IBM's convulsions in the plunging mainframe market would necessisitate at least adouble loop solution, e.g., by massive re-structuring, downsizing, and layoffs. Triple looplearning is relatively rare and is required only under extreme circumstances such as when acompany is facing a life-threatening trauma. The company may then have to re-invent itself byre-examining the corporate mission. Computervision and Wang Laboratories, for example, haveabandoned their proprietary hardware to become software and service providers. At thefunctional level, the IS function is also facing a critical challenge to its fundamental mission asusers increasingly declare their independence by developing their own localized solutions withinat departmental levels.

II. Becoming a Learning Organization

Organizational ineffectiveness will remain as long as people keep on trying to arrangeorganizations on the basis of the classical principles. In other words, if learning and changing areonly limited at the most to double loop learning. The networked corporation is a triple loopsolution, a bold departure from the hierarchy. In the networked corporation, the dominantmechanism for controlling activities between the organization and its customers, suppliers andpartners is the market.

The market is directed primarily towards the regulation of business relationships. In the lastdecade, -organizations have become increasingly connected through mergers, alliances, jointventures, franchises and partnerships. These new relationships result in business and socialnetworks that emphasize equality, cooperation, and collaboration.

Another fundamental development is the integration of information technology into thecorporate infrastructure as the medium of communication and coordination, and as a driver oforganizational change. The use of IT has resulted in both flattening the organization andincreased individualization. Class and status as well as the old hierarchical forms are dissolving.The new challenge ahead is therefore to organize cooperative processes. [WIERDSMA92, p145-148]

In organizing collaborative processes, thinking, deciding, doing, and reflecting will no longer beseparate, but will be integrated. Thinking can no longer be separated from doing. In short,organizational learning must take place at all levels and by all groups concurrently. This is thelearning organization, an organization in which people learn through cooperation and cooperatewhile learning.

To become a learning organization, the networked corporation also has to be aware of the sourcesof learning disabilities and take steps to overcome them. For a start, the organization can reviewits education programme. There are two forms - organizational education (in the narrow senseof education) and the organizational training. Traditionally, training courses and educationcourses are separate programmes aimed at different target groups. Lower and middle managerstend to be sent to training courses - learning to do. By contrast, higher and top managementenrol in courses that emphasize increasing knowledge and insight. This is education in thenarrower sense - learning to think.

For effective learning, education should be combined with training in one and the sameprogramme. [WIERDSMA92, p 901 IT implementation and planning in fact are processes thatprovide the convergence of thinking and action. They are therefore ideal learning tools. Otherlearning disabilities are culture related and may be corrected by changing the appraisal andreward systems, and emphasizing cooperative as opposed to competitive behaviors. Cognitivelimitations especially in tackling systems problems may be overcome by training in systemsthinking. Conscious application of the five disciplines can overcome most of the remainingimpediments to learning.

Finally, the success of the networked organization depends to a large extent on not just itsmanagement's ability, but also the employees' willingness and courage, to learn. Learningprovides the basis for adaptation and renewal in the face of rapidly changing relationships vis-a-vis competitors, partners, suppliers and customers. The opportunities afforded by IT processes,in particular IT strategic planning, will put management and lower level employees in a positionto reegineer processes, redefine relationships, and to learn together as a team.

ExecutionLearning Opportunities- communicating and sharing

knowledge- adapting to new system- reacting to and resolving

new problemsLearning Disabilities- competitive culture impedes

knowledge sharing- old reward system penalizes

new approaches- cultural limitations lead to

hiding of errors

Implementation

ConvergenceLearning Opportunities- integrating business with

technical knowledge- designing new process,

infrastructure and rewardreward system

- exploiting interdependence

Learning Disabilities- inability to perceive inter-

connectedness- lack of shared understanding- cognitive limitations in

identifying leverage points

JsageDivergenceLearning Opportunities- exploring new relationsips- exploring new work processes- evaluating new technology

Learning Disabilities- lack of courage to face new

challenges- incentive system discourages

innovations- cultural limitations lead to

rejection of new ideas

Feasibility

Assimilation StudyLearning Opportunities- building shared vision-identifying new opportunitiesfrom technology

Learning Disabilities- cognitive limitations in grasping

new opportunities and challenges- interpersonal limitations in team

learning

PlanningChart C.1: Learning Opportunities and Disabilities in IT processes

Learning Discipline

Personal Masterv

Mental Models

- Shared Vision

- Team Learning

- Systems Thinking

Learning Disabilities

Lack of courage to facenew challenges

Cultural limitations thatlead to rejection of newideas

Failure to grasp newopportunities

Knowledge sharing impededby competitive culture

Hiding Errors

Lack of shared understanding

Reward system that penalizesnew approaches

incentive system thatdiscourages innovations

Inability to identifyleverage points

inability to understand longterm impacts of decisions

Chart C.2: Learning Disciplines and Disabilities

IT Processes Learning Loops Results

Usage Single Loop Incremental improvementsNew agreements and rules

Implementation Single and Re-engineered work processesDoubleLoops New infrastructure

Planning Single, Double New missionand Triple Loops New relationships

New principlesChart C.3: Learning results in IT processes

Bibliography

[ABB-HBS91] William Taylor,Harvard Business Review, "The Logic of Global Business:An Interview with ABB's Percy Barnevik", March-April 1991[ACKERMAN92] Mark Ackerman, "Answer Garden : Building an OrganizationalMemory", Ph.D. Thesis Draft, Sloan School of Management, December 30, 1992.[ALLEN77] Thomas J. Allen, Managing the Flow of Technology, MIT Press, Cambridge, MA.[ARGYRIS77] Chris Argyris, "Double loop learning Organizations", Harvard BusinessReview, September-October 1977.[ARGYRIS78] Chris Argyris and Donald Schon, "Organizational learning : A Theory ofAction Perspective", Reading, MA: Addison-Wesley.[ARGYRIS91] Chris Argyris, "Teaching Smart People how to Learn", Harvard BusinessReview, May-June, 1991.[BARTLETT86] C.A. Bartlett, "Building and Managing the Transnational: The NewOrganizational Challenge", in M.E Porter (ed.), Competition in Global Industries, pp 367-401, Harvard Business School Press, Boston, MA, 1986.[BGOSHAL89] C.A. Bartlett and S. Goshal, Managing across borders, Harvard BusinessSchool Press, Boston MA, 1989.[BOYNTON92] Andrew C. Boynton, Gerry C. Jacobs & Robert W. Zmud, SloanManagement Review, "Whose Responsibility is IT Management?", by Summer 1992[BRAND87] Stewart Brand, The Media Lab: Inventing the Future at MIT, Viking Penguin,New York, 1987.[BULLEN88] Christine V. Bullen and Robert R. Johansen, "Groupware : A key toManaging business teams?", CISR Working Paper #169, Center for Information SystemsResearch, MIT, Cambridge, May 1988.[BWEEK,FEB93] "The Virtual Corporation", Business Week, February 8,1993.[BYTE93] Jeffrey Hsu and Tony Lockwood, "Collaborative Computing", BYTE, March1993.[CAMP89] Robert C. Camp, Benchmarking, ASQC Quality Press, Milwaukee, 1989.[COMWORLD93] Don Monaco, "The CIOs' new clothes", ComputerWorld, April 19,1993, Vol 27, No 16.[CARLSSON76] Barbara Carlsson, Peter Keane and J. Bruce Martin, "Learning andProblem solving : R&D organizations as Learning Systems", Sloan Management Review,Spring 1976.[CORP90] Michael S. Morton, The Corporation of the 1990s, Introduction, OxfordUniversity Press, New York, 1991.[CORNINC89] Corning, Inc., : "The Age of Hierarchy is Over", New York Times,September 24, 1989.[DAVENPORT92] Thomas H. Davenport, Robert G. Eccles and Laurence Prusak,"Information Politics", Sloan Management Review, Fall 1992.[DRUCKER88] Peter Drucker, "The Coming of the New Organization", HarvardBusiness Review, Jan-Feb 1988[DUNCAN79] R.B. Duncan and A. Weiss, "Organizational Learning; Implications fororganizational design", Research in Organizational Behaviour, 1979.[FELDMAN86] Feldman, M.S., "Constraints on communications and electronic mail",Proceedings, MCC Conference on Computer-supported Cooperative Work, Dec 1986, pp.73-90.[FIELDS] Tom Richman, "Mrs Fields secret ingredient", INC., Oct 1987[FORTUNE,MAY92] "The Search for the Organization of Tomorrow", Fortune, May 18,1992.[GEUS88] Arie P De Geus, "Planning as Learning", Harvard Business Review, Mar-April1988.

[HENDERSON891 John C. Henderson and N. Venkatraman, "Strategic Alignment: AProcess Model for Integrating Information Technology and Business Strategy", SloanSchool of Management Working Paper #3086-89, 1989.[HENDERSON89.11 John C. Henderson, "Building and Sustaining Partnership betweenLine and IS managers", Sloan School of Management Working Paper #3085-89, 1989.[IBM88] H. Jeff Smith, "IBM Computer Conferencing", Harvard Business School CaseStudy, 1988.[ISAACS92] William Isaacs and Peter Senge, "Overcoming limits to learning inComputer-based learning environments", European Journal of Operational Research,May 1992.[ISENBERG84] Daniel Isenberg, "How Senior Managers Think", Harvard BusinessReview, Nov-Dec 84.[ITAMI87] Hiroyuki Itami, Mobilizing Invisible Assets, Harvard University Press, BostonMA, 1987.[JOHANSEN88] Robert Johansen, Groupware: Computer Support for Business Teams, TheFree Press, New York, 1988.[KIESLER86] Sara Kiesler, "The hidden messages in computer networks", HarvardBusiness Review, Jan-Feb, 1986.[KIM90] Daniel H. Kim, "Total Quality and Systems Dynamics : Complementaryapproaches to Organizational Learning", MIT Sloan School of Management WorkingPaper #D-4162, 1990.[KIM92] Daniel H. Kim, Organizational Learning and Individual Learning: Where theTwain shall meet?", MIT Sloan School of Management Work paper #D4114-1, Sept 8,1992.[KOLB71] D.A. Kolb, "Individual Learning styles and the Learning Process." SloanSchool Working Paper 535-71, MIT, 1971.[LARSON89] Larson, Carl E., and LaFasto, Frank M.J., Teamwork, Sage Publications,Newbury Park, CA, 1989.[LBARTON92] Dorothy Leonard-Barton, "The Factory as a Learning Laboratory", SloanManagement Review, Fall 1992.[LEVY93] Steven Levy, CEO of Bolt Beranek and Newmen, Inc, during TQMpresentation, Feb 26, 1993.[McKINNON93] Interview with Richard Mckinnon, Sloan School of ManagementProfessor, February, 1993.[MEYER92] Arnoud De Meyer, "Management of International R&D operations", Chapter8, Technology Management and International Business, John Wiley & Sons Ltd, England,1992.[MITSUBISHI90] "Mighty Mitsubishi", Business Week, September 24, 1990.[NONAKA91] Ikujiro Nonaka, "The Knowledge-Creating Company", Harvard BusinessReview, November-December 1991.[PETERMK92] David E. Meen and Mark Keough, "Creating the learning organization",The McKinsey Quarterly, 1992 Number 1.[PETERS92] Tom Peters, Liberation Management, pp 50-52, Alfread A. Knopf, Inc., 1992.[PETERS93] Interview with Tom Peters, "A harsh indictment of CIOs", InformationWeek,February, 1993.[PRAHALAD90] C.K. Prahalad and G. Hamel, "The Core Competence of theCorporation", Harvard Business Review, May-June 1990.[ROCKART86] T. Gerrity and J.F. Rockart, "End-User Computing: Are you are Leader ora Laggard?", Sloan Management Review, Summer 1986, pp25-34.[ROCKART91] John F. Rockart and James E. Short, "The Networked Organization andthe Management of Interdependence", The Corporations of the 1990s, Chapter 7.[SCHEIN92] E. Schein, "Organizational Learning and Organizational Culture", SloanConvocation Seminar, Oct. 24, 1992.[SENGE90] Peter Senge, The Fifth Discipline,, DoubleDay, New York, 1990