an analytic approach for quantifying the value of e ......analytic tool. the methodology, the risk...

An analytic approachfor quantifying thevalue of e-businessinitiatives

by W. GreyK. KatirciogluS. BagchiD. Shi

G. GallegoD. SeyboldS. Stefanis

We describe an IBM strategic consultingoffering involving a methodology and ananalytic tool. The methodology, the Risk andOpportunity Assessment, provides asystematic approach for diagnosing problemsin the value chain of the enterprise, and forselecting and prioritizing e-business initiatives.Applying this methodology involves the use ofan analytic tool, the Value Chain ModelingTool, that uses management science andoperations research techniques, as well astechniques from the domains of finance andsupply chain management, to model the end-to-end value chain of the enterprise. Thisapproach has been successfully used toimprove the financial and operatingperformance of several enterprises.

Much of the value associated with e-business initi-atives comes not only from the improvements to in-formation technology (IT) infrastructure, but alsofrom business transformations that impact an orga-nization’s people and processes. A critical part of cre-ating business value is identifying the business pro-cesses to transform and selecting the right initiativesto enable the transformation. The focus on businessvalue continues during the transformation process,as decisions are made about project scope, scale, anddirection.1–3

To help customers realize greater business valuefrom their e-business investments, the IBM ResearchDivision teamed up with IBM Business ConsultingServices and developed a new business transformationmethodology—the Risk and Opportunity Assessment (or

Assessment, for short). This methodology provides asystematic approach for diagnosing problems in thevalue chain4 and for selecting and prioritizing e-busi-ness initiatives. It has been tailored to address issuesin a number of domains, including supply chain man-agement and product life-cycle management.5

The Assessment enables customers—under theguidance of IBM consultants—to identify solutionsthat improve business responsiveness and resilience.It also provides a consistent framework for evalu-ating and prioritizing e-business initiatives. Under-standing the impact of e-business initiatives on bus-iness performance is especially useful in today’sdifficult economic environment, because companiesare investing significant effort in developing businesscases to justify their e-business initiatives.6,7

The Assessment incorporates a sophisticated ana-lytic tool, the Value Chain Modeling Tool (or the tool,for short), that quantifies the relationship betweene-business investments and business value. This tool,which is implemented as a Microsoft Excel spread-sheet using Microsoft Visual Basic** macros, isequipped to evaluate scenarios involving risks anduncertainties in the customer’s business environ-ment.

The Value Chain Modeling Tool is designed to as-sess how changes in the speed, responsiveness, and

�Copyright 2003 by International Business Machines Corpora-tion. Copying in printed form for private use is permitted with-out payment of royalty provided that (1) each reproduction is donewithout alteration and (2) the Journal reference and IBM copy-right notice are included on the first page. The title and abstract,but no other portions, of this paper may be copied or distributedroyalty free without further permission by computer-based andother information-service systems. Permission to republish anyother portion of this paper must be obtained from the Editor.

GREY ET AL. 0018-8670/03/$5.00 © 2003 IBM IBM SYSTEMS JOURNAL, VOL 42, NO 3, 2003484

variability of business processes affect financial andoperational performance. The tool provides an in-tegrated representation of the client’s end-to-endvalue chain, from the initial design of new productsto after-sales support. It is thus especially effectivefor evaluating e-business initiatives focusing on sup-ply chain management, procurement, distributionand logistics, and product life-cycle management. Itcan be used to assess the impact of a broad array ofpotential transformations, ranging from targeted in-itiatives that affect only a single business process, tobroader initiatives that span functional and even en-terprise boundaries. Although the current versionof the tool has been tailored for clients in manufac-turing and distribution industries, the overall mod-eling approach and the Risk and Opportunity As-sessment methodology can be applied acrossindustries.

The analytic techniques used in the tool are basedon results from two key streams of applied research.The first results from an IBM Research effort thatuses management science and operations researchtechniques to model the impact of operational driv-ers on the operating performance of an enterprise.The second results from a cross-disciplinary IBM Re-search effort that integrates tools and techniquesfrom the domains of finance and supply chain man-agement to improve overall business performance.These two streams of research have been appliedbroadly to improve the financial and operating per-formance at a number of IBM business units, includ-ing Personal Computer, Microelectronics, and Stor-age.

The structure of the paper is as follows. First, we pro-vide a brief overview of the Risk and OpportunityAssessment methodology. Then we describe the ar-chitecture of the Value Chain Modeling Tool andits key components. Next, we describe the use of sce-narios and their role in the Assessment. We then dis-cuss a case study and illustrate the scope of our meth-odology. In the last section we present ourconclusions.

The Risk and Opportunity Assessment

The Risk and Opportunity Assessment is a meth-odology developed by IBM for diagnosing problemsin an enterprise value chain, and for pinpointing ap-propriate solutions. It uses an analytic tool, the ValueChain Modeling Tool, to help clients understand therelationship between financial and operational driv-ers and enterprise performance. As detailed below,

the Assessment involves scenario analysis in whichthe tool is used to determine the impact of proposedsolutions, to compare different options, and to as-sess their likely effectiveness within the context ofa particular business and operating environment.

The Risk and Opportunity Assessment follows athree-step process. The first step is data collection.During this stage of the engagement, the consultingteam works with the client to define data require-ments, to develop a data collection plan, and to carryout the data collection effort. The team also reviewsbackground information, documents assumptionsand sources, and monitors and validates the data col-lection process.

The second step of the engagement is modeling andanalysis using the Value Chain Modeling Tool. Thisstage involves building and validating a baselinemodel, defining and validating financial and oper-ational drivers, and developing the what-if scenar-ios used for evaluating proposed solutions. After thebaseline model and the scenarios have been created,the consulting team performs what-if analyses to de-termine the business impact of each proposed so-lution.

Figure 1 illustrates the use of the Value Chain Mod-eling Tool in a Risk and Opportunity Assessmentengagement. A data set representing a baselinemodel includes market characteristics, as well asvalue chain attributes. This information, along withwhat-if scenarios that represent potential changesto the value chain and its operating environment,serve as model inputs. The analytic model, embed-ded in the tool, processes this information and cal-culates the business impact of each what-if scenario.

The third and final step of the engagement is devel-oping a set of findings and recommendations. Dur-ing this stage, the team performs a high-level cost-benefit analysis, selects and prioritizes solutions, anddevelops an action plan for business transformation.The consulting team and the client agree on the stepsto be taken next, and develop a timetable for fol-low-on activities.

The Risk and Opportunity Assessment is particu-larly well suited for providing insight and guidanceto a company on the path to becoming an “on de-mand business.” An on demand business is an en-terprise whose business processes are flexible andresponsive to shifts in customer demand, new mar-ket opportunities, and external threats. Because the

IBM SYSTEMS JOURNAL, VOL 42, NO 3, 2003 GREY ET AL. 485

Assessment can help quantify the effect of improvedresponsiveness and flexibility at various points in thevalue chain, it can help identify business processesmost in need of improvement. It can also help as-sess the impact of increased business process inte-gration, both within the firm, and in its interactionswith customers and suppliers.

The Value Chain Modeling Tool

In this section, we discuss the Value Chain Model-ing Tool in greater depth. We begin with an over-view of its structure and information flows. Next, wedescribe its inputs, and the financial and operationalmetrics it computes. We then briefly discuss the tool’skey analytic components and describe our approachfor modeling the relationship between the main op-erational and financial drivers and enterprise per-formance measures. We conclude this section by de-scribing the use of the Value Chain Modeling Tooltogether with other modeling tools in order to ex-tend its modeling capability to include componentsat the business process, software application, and in-formation technology (IT) infrastructure levels.

Structure and information flows. The Value ChainModeling Tool has five main components: the datastore, the operational calculator, the financial cal-culator, the operational metrics calculator, and thefinancial metrics calculator.

The primary function of the data store is to struc-ture and preprocess the inputs to the Value ChainModeling Tool and to serve as an integration layerfor all other components. The operational calcula-tor uses operations research models to link changesin physical, temporal (delays), and informationalflows to changes in operational performance mea-sures. The financial calculator uses an integrated setof financial models and accounting identities to mapchanges in operational performance measures (ascalculated by the operational calculator) into changesin financial performance measures. The operationaland financial metric calculators compute the arrayof operational and financial metrics necessary to gen-erate financial statements.

Figure 2 shows the Value Chain Modeling Toolstructure and information flows. The tool inputs in-

Figure 1 Using the Value Chain Modeling Tool

BUSINESS IMPACT

• STRATEGIC METRICS• TACTICAL METRICS• FINANCIAL STATEMENTS• OPERATIONAL PERFORMANCE

WHAT-IF SCENARIOS

• CHANGES IN OPERATIONAL VALUE DRIVERS• CHANGES IN FINANCIAL VALUE DRIVERS• CHANGES IN RISK FACTORS

MARKET CHARACTERISTICS

• PRODUCT COMPLEXITY• DEMAND VARIABILITY• CUSTOMER SERVICE REQUIREMENTS• GEOGRAPHICAL DISPERSION• LEVEL OF COMPETITION

VALUE CHAIN ATTRIBUTES

• SPEED• FLEXIBILITY• EFFICIENCY• RESPONSIVENESS• INNOVATION

BASELINE MODEL DATASET

VALUE CHAIN MODELING TOOL

• OPERATIONAL MODELS• FINANCIAL MODELS

GREY ET AL. IBM SYSTEMS JOURNAL, VOL 42, NO 3, 2003486

clude value chain attributes and market character-istics as well as financial and operational drivers. Theinputs are preprocessed and stored by the data store(not shown in Figure 2). Operational data are trans-ferred (via the data store) to the operational calcu-lator, which calculates operational performance met-rics, such as inventory levels, throughput levels,product development cycle times, and network cy-cle times. These values are stored in the data store,and then passed to the financial calculator, along withadditional financial data. The financial calculatorproduces financial performance data, such as rev-enue, costs, cash flows, and balance sheet items.These data are then used by the financial metrics cal-culator to compute financial performance metricssuch as profit margins, asset utilization measures, andcash-to-cash cycle time. In an analogous fashion, out-puts of the operational calculator are passed to theoperational metrics calculator, which computes op-erating performance metrics such as order fill rates,delivery lead times, and service levels.

Inputs and outputs. In defining the tool inputs, ourobjective was to gather sufficiently detailed informa-

tion to enable credible modeling of value chain per-formance, while keeping data-gathering require-ments tractable. In order to accommodate a widerange of value chain issues, the model has a rela-tively broad set of inputs. However, for many anal-yses, only a subset of model inputs is required to gen-erate acceptable results. We describe below the keyinputs for the tool.

● The enterprise inputs consist of data relevant to theenterprise as a whole (the term “enterprise” mayrefer either to an entire business, or to a partic-ular business unit or division). The data includelogistics and other supply chain management costs,as well as high-level-income-statement, balance-sheet, and cash-flow-statement items.

● The product inputs consist of data about key prod-uct categories or product families. The data includeinformation on product sales, demand variability,customer requirements, order management andplanning cycle times, production characteristics,indirect spending, inventory costs, and productquality.

Figure 2 Value Chain Modeling Tool structure and information flows

FINANCIAL PERFORMANCE METRICS

OPERATIONAL PERFORMANCE METRICS

OPERATIONAL CALCULATOR

OPERATIONALMETRICS CALCULATOR

OPERATIONAL PERFORMANCE DATA

VALUE CHAIN ATTRIBUTES AND MARKET CHARACTERISTICS

FINANCIAL AND OPERATIONAL DRIVERS

FINANCIAL CALCULATOR

FINANCIALPERFORMANCE DATA

FINANCIALMETRICS CALCULATOR


● The supply inputs provide information about keyraw materials and components used in production.They include direct materials costs, supplier re-sponsiveness, component quality, and inventorycarrying costs.

● The product configuration matrix describes the high-level relationship between product categories andcomponent classes. It identifies the quantity of eachcomponent required to build each product.

● The distribution inputs characterize network flowsthrough a multi-echelon distribution network.They provide details on transportation and han-dling cycle times, as well as transportation and stor-age costs.

● The after-sales support inputs provide informationabout post-sales support, including customer ser-vice requirements, costs associated with warran-ties, product liability claims, recalls, and field ser-vice.

● The product development inputs give a detailed viewof product development costs and cycle times.Product development costs are attributed to anumber of activities, including design, prototyp-ing, mock-ups, data search, testing, tooling, andengineering changes. This information is furtherbroken down by stages in the product developmentprocess.

The tool generates a rich set of output metrics thatcharacterize financial and operating performance.There are five categories of outputs as described be-low.

1. The performance summary consists of key high-level financial and operational metrics. It includesa number of high-level strategic and tactical met-rics typically included in a balanced scorecard.9,10

2. The financial statements include a high-level in-come statement, a balance sheet, and a cash flowstatement.

3. The detailed metrics are a suite of financial andoperational metrics. They include profitability ra-tios, cash flow ratios, asset utilization metrics, andmeasures of operating efficiency. They also in-clude a broad array of metrics that assess supply-chain and product-development performance.

4. The product life-cycle management and product in-novation management metrics are a suite of finan-cial and operational metrics that characterize theeffectiveness of the product-development andproduct-innovation processes. They include de-tailed reporting of product development costs andcycle times, broken down by product developmentactivity and by stage in the product development

pipeline. They also include a number of impor-tant measures of product innovation effectiveness.

5. The metrics by product are a set of summary anddetailed financial and operational metrics brokendown by product category or family.

Operational and financial calculators. The opera-tional calculator uses stochastic quantitative mod-els from the field of operations research to linkchanges in physical, temporal and informationalflows to changes in operational performance.11–14 Theoperational calculator estimates the impact of valuechain characteristics, such as supply delays, produc-tion cycle times, and demand variability, on metricssuch as inventory, unit sales, and customer service(see sidebar for an illustration). It also calculates theimpact of product development characteristics, suchas product development delays, on unit sales.

Value chain processes that are modeled in detail in-clude planning, product development, supply man-agement, production, distribution, and order fulfill-ment. These processes affect value chainperformance by introducing delays, variability, andconstraints into the system. Delays make the enter-prise less responsive to changes in the external bus-iness environment. Variability influences the predict-ability of value chain performance, making it harderto meet customer service targets. Constraints, suchas limited production capacity, affect the responsive-ness to change of the enterprise.

More specifically, companies with long product de-velopment cycle times may be late to market, andtherefore may lose market share to more agile com-petitors. Companies with long planning cycle timesare less responsive to variability in customer demand,and are thus forced to compensate by either hold-ing more inventory, or by sacrificing delivery perfor-mance. Poor planning processes can introduce ad-ditional system variability, as happens when, forinstance, poor demand planning results in inaccu-rate demand forecasts. This makes it harder for acompany to effectively match production to customerdemand, causing poor supply chain performance. Fi-nally, manufacturing capacity constraints and sup-ply constraints limit a company’s ability to respondto major upturns in customer demand.

The financial calculator uses accounting identitiesand financial models to map the output of the op-erational calculators from physical units to financialmeasures.15,16 Product development costs, produc-tion costs, logistics costs, and inventory costs are


Quantifying the financial impact of operational changeQuantifying the financial impact of operational change

I n order to clarify the role of analytic models in the Value Chain Modeling Tool, we describe here a simplified analytic model that illustrates our approach for quantifying the financial impact of changes to operational performance.

Consider a manufacturing company with a planning process that determines the products to manufacture. This process begins with customer demand forecasts provided by the marketing and sales organization. Next comes master planning, performed jointly with the manufacturing organization, which is an assessment of the manufacturing capability. Following the completion of the planning process, the company manufactures its products and ships them to the final customer, usually through a distribution network comprised of multiple distribution centers and warehouses.

Suppose this entire process—from creating the demand forecast, to actually shipping products to the final customer—takes 32 days. This time lag, which we refer to as “end-to-end cycle time,” introduces uncertainty in the planning process. In essence, when the company creates its demand forecast, it needs to predict demand 32 days in the future—the time when customers will actually be purchasing its products. Because customer demand in the marketplace is always changing, the further ahead a company needs to predict demand, the less accurate its forecasts will be. To compensate, the company needs to hold more inventory.

To compute the impact that a reduction in planning cycle times has on inventory levels, we begin by using a simple heuristic to estimate the impact of cycle time delays on forecast accuracy. We then use a simple safety stock calculation to estimate the impact of the resulting forecast accuracy on inventory levels.

One rule of thumb uses the square-root function to calculate forecast error as a function of end-to-end cycle time. Suppose the company can forecast daily demand a week in advance with 30 percent error. Then the error of forecasting daily demand 32 days in advance is (32/7)½(30%) = 64 percent. This error determines the inventory required to accommodate the uncertainty, known as safety stock. We estimate this quantity using a standard safety stock calculation.1 In a stationary environment the safety stock is given by a multiplier k�, where k is the safety factor, determined by the service target of the company, and � is the standard deviation of demand during lead time (32 days). Suppose daily demand can be represented as independent and identically distributed random variables with a mean of 100 units per day. Then, � = (�i=1,..,32(100 (i/7)½(30%))2)½ = 261. It is assumed the standard deviation of demand i days ahead is 100(i/7)½(30%). Suppose the objective is to meet demand with off-the-shelf inventory 98 percent of the time. This gives us a safety factor of k = 2.05 (k is simply the independent variable value corresponding to a 0.98 value for the cumulative standard normal distribution). Therefore, the company has to hold a safety stock of k� = 2.05x261 = 535 units (or 5.35 days of supply).

Now, suppose the company is evaluating a business transformation that will reduce total planning cycle time by ten days. The end-to-end cycle time will be 22 days instead of 32 days. Then, � = (�i=1,..,22(100 (i/7)½(30%))2)½ = 180. Therefore, the safety stock will be as follows: k� = 2.05x180 = 369 (or 3.69 days of supply). Hence, by implementing this business transformation, the company can cut inventory by about 166 units (over 1.5 days of supply).

I

Reference

1. A. C. Hax and D. Candea, Production and Inventory Management, Prentice-Hall, New York (1984)


modeled in considerable detail. There are also cal-culations to estimate costs associated with process-ing returns, costs due to defective products, and op-portunity costs associated with lost sales. A simplifiedanalytic model of this type is illustrated in the side-bar.

The financial calculator employs a “bottom-up”modeling approach whereby low level metrics areaggregated into higher level metrics. In general a

firm’s cost structure is modeled in terms of both fixedand variable costs. For example, manufacturing costsare decomposed into fixed and variable productioncosts. Variable production costs are further decom-posed into direct materials costs and costs associ-ated with the procurement and delivery of direct ma-terials. Different stages of production are modeledto accurately reflect the incremental value associatedwith each step. This improves the accuracy of esti-mates of inventory values and cost of sales. Productdevelopment expense is itemized with over 20 ex-pense categories covering five stages in the productdevelopment cycle. The impact of cost drivers, suchas the number of product development projects andnumber of engineering changes, is modeled in de-tail.

Using the Value Chain Modeling Tool with other e-business modeling tools. In this section, we posi-tion the Value Chain Modeling Tool with respect toother e-business modeling tools. In particular, we fo-cus on tools used to assess changes at the businessprocess, software application, and IT infrastructurelevels.

The IT infrastructure, the applications, and the bus-iness processes in an enterprise all play a critical rolein creating business value. IT infrastructure capabil-ities—including speed, flexibility, capacity, efficiency,resilience, and security—determine the types of ap-plications that can be run and their performance. Ap-plications support business processes in various func-

tional areas of the organization. These applicationsaffect the accuracy, speed, and productivity of thesebusiness processes. Business processes, in turn, af-fect operating performance at different points in thevalue chain, and these have a direct impact on theoverall business performance of the enterprise.17–20

Figure 3 illustrates the way these elements interact,that is, each one impacts the one downstream. TheIT infrastructure, for example, does not create bus-iness value directly; it does so through its role as anenabler of applications and business processes. Mod-eling the business impact of infrastructure invest-ments is thus a multi-step process. First, we analyzehow the proposed investment will affect infrastruc-ture performance. Then we analyze the downstreamimpact of changes in infrastructure performance—first on applications and business processes—thenon operational and business performance.

There are a number of analytic approaches for mod-eling the performance of IT infrastructure, applica-tions, and business processes.21 System performancemodeling uses techniques such as benchmarking,queuing theory, and simulation to evaluate the im-pact of changes to IT infrastructure on the perfor-mance of applications.22–24 The coverage of thesetechniques is illustrated in Figure 3 by the box la-beled SYSTEM PERFORMANCE MODELING. Consult-ing organizations apply best-practice methodologiesto redesign business processes, often in conjunctionwith the deployment of IT applications.25–27 Otherbusiness process analysis techniques, including bus-iness process simulations, can be used to model howchanges to business processes and IT applications af-fect operational performance.28,29 These techniquesare illustrated in Figure 3 by the box labeledBUSINESS PROCESS ANALYSIS.

Although these analytic approaches can be used tomodel some aspects of the value chain, they do notalways quantify the broader financial implications ofe-business investments.30,31 The Value Chain Mod-eling Tool, while designed to quantify financial im-pact, does not explicitly model the performance im-pact of investments in e-business infrastructure andapplications. As Figure 3 illustrates, integrating theoutputs of these lower-level models with the ValueChain Modeling Tool should lead to more accurateresults and a better understanding of business im-pact.

As an example, consider a system performance mod-eling effort that assesses the impact of an infrastruc-

Using the Value ChainModeling Tool in conjunction with

IT modeling tools and businessprocess analysis tools

leads to a better understandingof the business model.


ture investment on low-level IT metrics, such as re-sponse time and availability. Business processanalysis can then be used to estimate how this wouldaffect low-level operating metrics, such as processcycle times, rework rates, and throughput. Finally,this information can be passed to the Value ChainModeling Tool and used to assess business impact.

Analysis of what-if scenarios

An important function of the Value Chain Model-ing Tool is the use of what-if scenarios for evaluat-ing the benefits and risks of potential e-business in-itiatives. The tool offers a number of differentanalyses, including a financial impact assessment, acost-benefit analysis, a strategic fit analysis, an as-sessment of key performance indicators, and a sen-sitivity analysis. Examples of several of these anal-yses are provided in the brief case study discussedin the next section.

The tool supports two types of what-if scenarios: so-lution scenarios and risk scenarios. Solution scenar-ios can be used to characterize a broad array of po-tential initiatives, ranging from an end-to-end valuechain transformation to the deployment of a pointsolution. Each solution scenario identifies the op-erational metrics directly affected by the proposede-business initiative, as well as the expected impactof the transformation on the operational metrics. Forexample, a solution scenario designed to assess theimpact of an e-business initiative to support customercollaboration might specify the expected improve-ments in operational metrics such as forecast accu-racy and demand planning cycle time.

Risk scenarios assess the performance of proposedsolutions in a dynamic and volatile business environ-

ment. Risk scenarios specify changes in internal andexternal metrics associated with risky future businessevents, such as major shifts in customer demand orunexpected supply shortages. As an example, con-sider a risk scenario designed to assess the impactof a sudden decline in customer demand. Unit saleswould probably decline, and heightened industrycompetition would probably lead to a decrease inaverage selling prices. If the demand shortfall werecaused by financial difficulties in a particular cus-tomer segment, customer payments would mostlikely be delayed. Customers would probably becomemore demanding and might only make purchases ifa product were available within the desired lead time.

Solution scenarios provide a structured approach forquantifying the impact of initiatives on a firm’s fi-nancial and operational performance. Risk scenar-ios assess the responsiveness and robustness of theclient’s value chain in the face of an uncertain bus-iness environment. By analyzing pair-wise combina-tions of risk and solution scenarios, the impact ofproposed initiatives on the firm’s risk profile can beassessed.

Standardized sets of solution and risk scenarios canbe stored in a scenario library. Individuals knowledge-able in a particular domain can develop basic sce-narios that identify key operational drivers affectedby the solution. These scenarios can then be distrib-uted to other practitioners for their use. During anengagement, these standard scenarios serve as tem-plates that can be customized to reflect the charac-teristics of a specific client. Practitioners can also usethe scenario libraries to create new scenarios thataddress specific client needs.

Figure 3 Modeling tools for analyzing e-business performance

APPLICATIONS BUSINESS PROCESSES

IT INFRASTRUCTURE

OPERATIONAL PERFORMANCE

BUSINESS IMPACTBUSINESS IMPACT

SYSTEM PERFORMANCE MODELING• BENCHMARKS• QUEUING THEORY• SIMULATION

BUSINESS PROCESS ANALYSIS• BEST PRACTICE MODELS• ANALYTICAL MODELS

VALUE CHAIN MODELING TOOL


Case study

In this section, we use a case study to illustrate theapplication of the Risk and Opportunity Assessmentin practice. The case study shows how the ValueChain Modeling Tool can be used to analyze andprioritize a set of potential initiatives being consid-ered by the Motorcycle Division of a major automo-bile manufacturer. Although the study is based onwork performed with an actual client, the informa-tion we present has been altered to preserve con-fidentiality.

The Motorcycle Division produces a broad array ofon-road and off-road vehicles, including motorcycles,all-terrain vehicles, and motor scooters. The com-pany has a truly global reach and serves major cus-tomer markets in Europe, North America, Japan,and Southeast Asia. Its business is highly compet-itive, and the firm believes that its success dependson meeting a number of critical business objectives,

including increased product development effective-ness, improved responsiveness to market needs, in-creased cost competitiveness, and enhanced brandimage. In addition to its highly competitive industrystructure, a number of factors make the motorcyclebusiness particularly risky. These include high de-mand volatility, worldwide constraints on productioncapacity, and sensitivity to foreign exchange rate vol-atility.

The management of the Motorcycle Division wasconsidering fundamental changes in their businesspractices and asked IBM to help them assess a set ofnew initiatives designed to transform their valuechain. This transformation would address severalcritical “pain” points affecting competitiveness, in-cluding excessive inventory levels, high transporta-tion costs, and long manufacturing lead times. Ananalysis of the client’s value chain revealed a num-ber of underlying factors that required improvement,

Figure 4 The scenario for the demand planning solution (screen shot)


such as long production and planning lead times andpoor demand forecast accuracy. The client alsoneeded to streamline its complex distribution net-work and reduce excessive warehouse transferscaused by poor inventory positioning.

IBM worked with the client to identify a number ofspecific initiatives that could affect key value driversand thus lead to achieving the client’s business ob-jectives.

Two potential solutions that we analyze here in de-tail are Demand Planning and Supply Network Plan-ning. We also consider a third solution consisting ofthe simultaneous deployment of Demand Planningand Supply Network Planning. This solution had thepotential to deliver the complementary benefits ofthe first two solutions, while significantly reducingdeployment costs.

We decided to use the Value Chain Modeling Toolto quantify the business value associated with eachof these initiatives. We began by gathering the datanecessary to build and validate the base case (thebaseline model). After validating the base case withthe help of the client, we defined and validated thefinancial and operational drivers affected by each ofthe initiatives. We determined that the proposed De-mand Planning solution would improve demandforecast accuracy. It also would reduce demand plan-ning cycle times and order processing delays andwould decrease required inventory safety stock lev-els at the client’s distribution centers. We further de-termined that the proposed Supply Network Plan-ning solution would reduce procurement andproduction planning cycle times, and improve logis-tics efficiency. It also would enable the client to makefurther reductions in inventory safety stock levels,in addition to those supported by Demand Planning.

We then defined a set of risk and solution scenarios.A solution scenario was created for each proposedinitiative. Figure 4 shows the solution scenario forDemand Planning. For each of the operational driv-ers affected by the initiative, anticipated changes areentered. As shown in the figure, after the transfor-mation the demand planning cycle time is 14 days,instead of 30 days. Average order processing timesare anticipated to decline from 8.5 days to three days.Improved forecasting processes are expected to re-duce forecast error from 35 percent to 20 percent.Furthermore, more reliable planning is expected toreduce minimum required inventory safety stock lev-els from 30 days to 22.5 days. Two risk scenarios were

also created. In the first, worsening economic con-ditions lead to a sharp and unexpected decline in cus-tomer demand. In the second, an improving busi-ness climate results in a sudden increase in customerdemand.

After defining the scenarios, we performed a what-ifanalysis to assess the business impact of each of theproposed initiatives. Sample results of this analysisare shown in the remaining figures. Figure 5 showsthe annual financial benefit of each solution. The De-mand Planning solution has a higher benefit thanSupply Network Planning, and the combined solu-tion (Demand Planning and Supply Network Plan-ning) has an even larger benefit. For all solutions,the most significant savings are associated with re-ductions in inventory. These include lower inventorycarrying costs, and smaller inventory write-downs andwrite-offs. There is also a comparatively small (butsignificant) logistics cost savings associated with theSupply Network Planning solution.

We next performed a strategic fit assessment. Wefirst identified a set of Key Performance Indicators(KPIs) that were closely aligned with the client’s bus-iness objectives. The Value Chain Modeling Toolwas then used to assess how each proposed solutionaffected each of the KPIs. As shown in Figure 6, aSupply Network Planning deployment leads to amoderate improvement in Shareholder Value Addedand Return on Net Assets, and a large improvement

Figure 5 Financial impact of proposed solutions

LOGISTICS COST SAVINGSINVENTORY PRICE DECLINE SAVINGS INVENTORY CARRYING COST SAVINGS

DEMAND PLANNINGPLUS SUPPLY NETWORK PLANNING

DEMANDPLANNING

SUPPLY NETWORK PLANNING


in Cash to Cash Cycle Time and Inventory Turnover.Its impact on other metrics is negligible. A DemandPlanning deployment, on the other hand, has a stron-ger and broader impact. It leads to a large improve-ment in Shareholder Value Added, Cash to Cash Cy-cle Time, Inventory Turnover and Order to DeliveryLead Time. It also has a moderate impact on Re-turn on Net Assets, and On Time Delivery.

Finally, we performed a high-level cost-benefit anal-ysis—the results are shown in Figure 7.32 Implemen-tation costs are the same for Demand Planning andSupply Network Planning, but Demand Planning of-fers significantly higher business value. The com-bined Demand-Planning and Supply-Network-Plan-ning initiative is somewhat more attractive than

Supply Network Planning alone, however, becausecertain resources could be shared during a joint de-ployment. This creates an opportunity for additionalcost savings, making the combined Demand-Plan-ning and Supply-Network-Planning initiative an at-tractive option.

We used the risk scenarios defined earlier to test therobustness of each initiative to changes in the ex-ternal business environment. All initiatives provedrobust to risk; under both pessimistic and optimisticscenarios, they continued to have a positive impacton operational and financial performance. Scenarioanalysis also helped to assess how each initiative af-fected the firm’s resilience to business uncertainty.The Demand Planning initiative was particularlyeffective at increasing value chain resilience becauseit increased forecast accuracy and responsiveness.In the face of an unexpected decline in customer de-mand, the client would be able to more rapidly ad-just inventories to match lower levels of customerdemand. On the other hand, if demand suddenly in-creased, the Demand Planning initiative would im-prove the effectiveness of inventory planning. Thiswould reduce lost sales due to inventory stock-outs.

Summary

In this article, we describe a new strategy offeringand a decision-support tool designed to help IBM cus-tomers realize greater business value from their e-business investments. The strategy offering, the Riskand Opportunity Assessment, provides a systematicapproach for evaluating e-business initiatives. It moreclosely aligns e-business implementations with bus-

Figure 6 Results of the strategic fit assessment

Key Performance IndicatorsShareholder Value Added Operating IncomeReturn on Net AssetsCash to Cash Cycle TimeOperating Profit MarginInventory TurnoverDays Receivables OutstandingDays PayablesOn Time DeliveryOrder to Delivery Lead Time

From

-10%-1%1%

10%

To-10%

-1%1%

10%

Supply Network Planning

DemandPlanning

SNP and DP

Key

Figure 7 Results of the cost-benefit analysis

BASE CASE

IMPLEMENTATION COST

BU

SIN

ES

S V

AL

UE

DEMAND PLANNINGSUPPLY NETWORK PLANNING

DPSNP

==

DP

SNP

DP+SNP


iness strategy, and helps identify solutions that im-prove business responsiveness and resilience. It hasbeen customized to address a number of importantdomains, including supply chain management andproduct life-cycle management.

The Risk and Opportunity Assessment methodol-ogy uses an analytic tool, the Value Chain Model-ing Tool, to quantify the business value of e-busi-ness investments. The Value Chain Modeling Tooluses stochastic quantitative models from the field ofoperations research to assess how changes in valuechain speed, responsiveness and variability affect op-erational performance. The tool uses financial mod-els to evaluate the financial impact of changes in op-erational performance. The tool provides anintegrated representation of an end-to-end valuechain. Key processes that are modeled include prod-uct development, procurement, manufacturing, dis-tribution and logistics, sales and order fulfillment,and after-sales support. Although the tool has beentailored for manufacturing and distribution indus-try value chains, the modeling approach can be ap-plied across industries.

The Value Chain Modeling Tool is also equippedto evaluate the benefits and risks of potential e-busi-ness initiatives. The tool supports two types of what-ifscenarios: solution scenarios and risk scenarios. So-lution scenarios provide a structured approach forquantifying the impact of initiatives on a firm’s fi-nancial and operational performance. Risk scenar-ios assess the responsiveness and robustness of thevalue chain in the face of uncertainty in the businessenvironment. The tool can perform various analy-ses, including a financial impact assessment, cost-benefit analysis, strategic-fit analysis, and sensitivityanalysis.

Faced with a difficult economy and heightened com-petition, IBM customers are devoting greater re-sources to developing business cases to justify theire-business initiatives.33,34 Because the tool describedin this paper provides a rigorous, quantitative frame-work for assessing business impact, it is very usefulfor developing a business case. The tool has provenits effectiveness in defining e-business strategy andmanaging e-business initiatives in multiple client en-gagements.

Cited references and notes

1. J. Dempsey et al., “A Hard and Soft Look at IT Investments,”The McKinsey Quarterly 1, 126–137 (1998).

2. M. J. Davern et al., “Discovering Potential and RealizingValue from Information Technology Investments,” Journalof Management Information Systems 16, No. 4, 121–143 (2000).

3. J. Kersnar, ROI: The Age of Reason, CFO.com (July 2002);http://www.cfo.com/cfo_home.

4. M. E. Porter, Competitive Advantage: Creating and Sustain-ing Superior Performance, The Free Press, New York (1985).

5. Current offerings include Value Chain Services, Extended En-terprise Collaboration, and Product Lifecycle Management(PLM) Core.

6. C. Murphy and M. Kolbasuk McGee, “Running The Num-bers,” Information Week (September 30, 2002), pp. 28–36.

7. D. Rezendes Khirallah, “Services Market in Slump, DespiteBig Deals,” Information Week (September 23, 2002), p. 24.

8. G. Lin, M. Ettl, S. Buckley, S. Bagchi, D. D. Yao, B. L. Nac-carato, R. Allan, K. Kim, L. Koenig, “Extended-EnterpriseSupply-Chain Management at IBM Personal Systems GroupAnd Other Divisions,” Interfaces 30, No. 1, 7–25 (January–February 2000).

9. R. S. Kaplan, D. P. Norton, “The Balanced Scorecard—Mea-sures That Drive Performance,” Harvard Business Review 70,No. 1, 71–79 (January–February 1992).

10. R. S. Kaplan, D. P. Norton, “Using the Balanced Scorecardas a Strategic Management System,” Harvard Business Re-view 74, No. 1, 75–85 (January–February 1996).

11. A. C. Hax and D. Candea, Production and Inventory Man-agement, Prentice-Hall, New York (1984).

12. E. A. Silver, D. F. Pyke, and R. Peterson, Inventory Manage-ment and Production Planning and Scheduling (Third Edition),John Wiley & Sons, New York (1998).

13. S. Nahmias, Production and Operations Analysis, McGraw Hill,New York (2001).

14. C. F. Daganzo, Logistics Systems Analysis (Third Edition),Springer-Verlag, Heidelberg, Germany (1999).

15. R. A. Brealey and S. C. Myers, Principles of Corporate Finance,McGraw-Hill/Irwin, New York (2002).

16. A. Damodaran, Corporate Finance: Theory and Practice, JohnWiley & Sons, New York (1997).

17. C. Sylla and H. J. Wen, “A Conceptual Framework for Eval-uation of Information Technology Investments,” InternationalJournal of Technology Management 24, Nos. 2/3, 236–261(2002).

18. P. Tallon, K. Kraemer, and V. Gurbaxani, “Executive’s Per-ceptions of the Business Value of Information Technology:A Process-Oriented Approach,” Journal of Management In-formation Systems 16, No. 4, 145–173 (2000).

19. J. Ross and C. Beath, “Beyond the Business Case: StrategicIT Investments,” MIT Sloan School of Management (2001,unpublished).

20. C. Orlowski and E. Szczerbicki, “Evaluation of InformationTechnology Projects,” Cybernetics and Systems: An Interna-tional Journal 33, No. 6, 659–673 (2002).

21. T. Mayor, “Value Made Visible,” CIO Magazine 13, No. 14(May 1, 2000).

22. R. Jain, The Art of Computer Systems Performance Analysis,Wiley, New York (1991).

23. J. S. Bozman and R. Perry, Server Cost of Ownership in ERMCustomer Sites: A Total Cost of Ownership Study, White Pa-per, IDC, Framingham, MA (September 2001).

24. Value Proposition for E-Infrastructures: Cost/Benefit Case forIBM eServer, Management Brief, International TechnologyGroup, Los Altos, CA (May 2002).

25. T. Piper and C. J. Hoffman, Making ‘Req to Check’ a Reality,White Paper, IDC, Framingham, MA (October 2001).

26. P. Mitchell and R. Shah, Indirect Procurement Field Review:


Setting the Record Straight, AMR Research, Boston, MA (Au-gust 2001).

27. B. Robertson, E-Business: Impact on Infrastructure, MetaGroup, Stamford, CT (August 29, 2001).

28. E. Bhaskar, H. S. Lee, A. Levas, R. Petrakian, F. Tsai, andB. Tulskie, “Analyzing and Re-engineering Business Pro-cesses using Simulation,” Proceedings of the 1994 Winter Sim-ulation Conference (WSC 1994), ACM, New York (1994).

29. S. Bagchi, S. J. Buckley, M. Ettl, and G. Lin, “ExperienceUsing the IBM Supply Chain Simulator,” Proceedings of the1998 Winter Simulation Conference (WSC 1998), ACM, NewYork (1998).

30. L. Siegele, “How about now? A Survey of the Real-TimeEconomy,” The Economist (February 2, 2002), pp. 3–18.

31. M. Bradford and D. Roberts, “Does Your ERP System Mea-sure Up?” Engineering Management Review 30, No. 2, 20–22(2002).

32. In assessing benefits, a number of metrics were analyzed, in-cluding the financial benefits discussed earlier. In the exam-ple shown here, we show an index of business value. This in-dex was constructed by taking a weighted average of the KPIsused to assess strategic fit.

33. R. J. Bowman, “Fast Deployment, Quick Return Help SellIT in a Slow Economy,” Global Logistics and Supply ChainStrategies (October, 2001).

34. R. Banham and H. Rosenberg, ROI: Mad to Measure, CFO.com (September 2001); http://www.cfo.com/cfo_home.

Accepted for publication March 3, 2003.

William Grey IBM Research Division, Thomas J. Watson ResearchCenter, P. O. Box 218, Yorktown Heights, New York 10598([email protected]). William Grey is Program Director for Valueand Risk Modeling at the IBM Thomas J. Watson Research Cen-ter, in Yorktown Heights, NY. He is currently leading a team thatis developing analytic tools and methodologies which support newstrategic consulting offerings that quantify the business impact ofinformation technology investments. His research activities andinterests include applications of finance and accounting techniquesto e-business, supply chain management, risk management, andthe valuation of intangibles.

Kaan Katircioglu IBM Research Division, Thomas J. Watson Re-search Center, P. O. Box 218, Yorktown Heights, New York 10598([email protected]). Dr. Katircioglu is one of the leading re-searchers in supply chain management at the IBM Corporation.He has a B.S. degree in industrial engineering and an M.S. de-gree in statistics. He completed his Ph.D. in management scienceat the University of British Columbia in 1996. He joined IBM asa researcher at Thomas J. Watson Research Center the same year.Since then, he has worked on several projects for various divi-sions of IBM and its customers, published papers, and made nu-merous conference presentations. He has a number of inventionswith patents pending. He is a member of INFORMS, and IEEE.He serves as a committee member on Semiconductor Factory Au-tomation (SFA) within the IEEE Robotics and Automation So-ciety.

Sugato Bagchi IBM Research Division, Thomas J. Watson Re-search Center, P. O. Box 218, Yorktown Heights, New York 10598([email protected]). Dr. Sugato Bagchi is a research staff mem-ber at the IBM Thomas J. Watson Research Center in YorktownHeights, NY. His current research interest is in valuation of in-formation technology applications and infrastructure. He is work-

ing on the development and application of tools and techniquesthat help potential adopters of emerging technologies quantifythe predicted benefits in financial terms. He has participated innumerous strategy consulting engagements with teams of e-busi-ness strategy consultants. Previously, Dr. Bagchi worked on knowl-edge representation frameworks and methodologies for businessstrategy formulation and analysis. He also worked on the designand implementation of a business process modeling and simu-lation tool, as well as a specialization of this tool for multi-en-terprise supply chains. Dr. Bagchi has a Ph.D. degree in Elec-trical and Computer Engineering from Vanderbilt University,where he wrote a thesis on task planning under uncertainty.

Dailun Shi IBM Research Division, Thomas J. Watson ResearchCenter, 19 Skyline Drive, Hawthorne, NY 10532 ([email protected]). Dr. Shi is a research staff member in the SystemsAnalysis and Optimization group. He has a Ph.D. degree in man-agement, an M.B.A. in finance, and an M.S. in industrial engi-neering, all from Georgia Institute of Technology. He also hasan M.S. in mathematics from Brown University. Before his ca-reer in research at IBM, Dr. Shi worked as a manager at Citi-group with the responsibility for Network Crisis Management.His current primary research interest is in modeling techniquesapplied to problems in business, management, and technology.His specific research areas include risk management, intelligencefor business contracts, operations management, supply chain man-agement, outsourcing, and e-commerce.

Guillermo Gallego Industrial Engineering and Operations Re-search, Columbia University, 324 Mudd Bldg, 500 West 120th Street,New York, NY 10027 ([email protected]). Dr. Gallego is Pro-fessor and Chairman of the Industrial Engineering and Opera-tions Research Department at Columbia University. He receivedhis Ph.D. degree in 1988 from the School of Operations Researchand Industrial Engineering at Cornell University. Dr. Gallego’sresearch interests span several applied areas including produc-tion planning, inventory control, revenue management and sup-ply chain management. His work has been supported by govern-ment and industrial grants. His graduate students are associatedwith prestigious universities.

Dave Seybold IBM Global Services, 100 East Pratt Street, Bal-timore, MD 21202 ([email protected]). Mr. Seybold is theleader of IBM Global Services Product Lifecycle Managementand Innovation consulting services. He has 14 years of experi-ence in business and technology consulting and is skilled in theareas of innovation and product development, supply chain man-agement, and information technology architecture. Prior to hiscurrent responsibility, he led IBM Global Supply Chain Practice.Mr. Seybold holds degrees in quantitative business analysis andeconomics from Pennsylvania State University, an M.S. in op-erations research, and an M.B.A. from the University of Mary-land Smith School of Business.

Stavros Stefanis IBM Global Services, 136 Crawley Falls Rd,Brentwood, NH 03833 ([email protected]). Dr. Stefanis is aPrincipal, Business Consulting Services, currently responsible forSupply Chain Management Competency development in IBMGlobal Services. Dr. Stefanis was the director of the solutions ar-chitecture group in Syncra Systems. He has been a solutions ar-chitect for IBM’s Supply Chain Optimization Practice and wasa senior architect for i2 Technologies. He earned his doctorateat the Centre for Process Systems Engineering at Imperial Col-lege in London, where his research focused on applied opera-


tions research and process systems engineering. He has publishedseveral papers in this area including a chapter on life-cycle de-sign in Webster’s Encyclopedia. He has nearly 14 years of softwaredevelopment and consulting experience in the supply chain op-timization area, including significant time dedicated to processmodeling, research and development. His clients have includedUnilever, BP, Philips Consumer Electronics, Philips Lighting,Frito-Lay, PepsiCo, World Wide Retail Exchange, E2open, Mo-torola, Honda, Southwire, Yazaki, and many others.


an analytic approach for quantifying the value of e ......analytic tool. the methodology, the risk...

Documents