the data-driven credit union: powering transformation with … · 2018. 5. 31. · the data-driven...
TRANSCRIPT
The Data-Driven Credit Union: Powering Transformation with
Advanced Analytics April 20, 2018
Adoption and execution
Workflow integration
Modeling insights
Dataecosystem
Sourceof value
AdvantEdge AnalyticsAreas of Differentiation
BusinessConsultingServices
Insights Activation
Reporting andPerformanceManagement
PredictiveAnalyticsSolutions
DataManagement
Data Transformation Strategy
Integrated AdvantEdge Analytics E2E Solutions and Services
Areas of differentiation for most analytics and tech vendors
Our Value Proposition and Differentiation
1
Growing Team + Endorsements
Key Endorsements
Growing Team130 + AdvantEdge
employees
25 + Data Scientists20+ Data Engineers
20 + Data Translators
+ 115 clients from Acquisition of
Sequan
Growing Client List
Growing Clients, Team and Endorsements
2
Before we begin – a quick poll
3
Spectrum of Data and Analytics
Advanced Analytics
• Prescriptive AnalyticsWhat is the best next action for us to take?
• Predictive AnalysisWhat will happen next?
10
9
Guided Analytics
• ForecastWhat if these trends continue?
• Statistical AnalysisWhy is it happening?
8
7
Reporting/Self Service
• AlertsWhat are best actions?
• Selective drill downWhere is the problem?
• Ad Hoc Queries How many, how often, where?
• Standard Reports What has happened?
6
5
4
3
Data• Clean Data
• Raw Data21
4
5
30%30%
50%50%
90%90%
What % of all data in the world was created in the last 2 years?What % of all data in the world was created in the last 2 years?
Data and Analytics is a key priority for Credit Unions…
Common business priorities
Grow member base
Deliver a best in-class digital member experiences
Optimize risk and manage losses
Grow number of products per member and wallet share
73%of credit unions see analytics as a way to significantly transform the way they do business
SOURCE: CUNA 2016 credit union member survey on data and analytics 6
…but most credit unions have been unable to drive business value
Have business-driven analytic initiatives26%
Have a comprehensive front-line adoption approach9%
72% Indicate most of their member data is NOT easily accessible
SOURCE: CUNA 2016 credit union member survey on data and analytics 7
The case for change
Driving analytics led transformation
8
Contents
2018 ≠ 1980Costs of data storageand processing
Dataavailability
Math
20181980 1950’s 1980’s 2010’s
DeepLearning A branch of ML
Machine LearningA major approach to realize AI
Artificial Intelligence The science of making intelligent machines
Basic demo-graphic data (e.g.,city, income)
Trans-actions data (e.g., ATMs, mobile-apps)
Gov. agencies (e.g., tax payment report, updated demo-graphic data)
Regular survey / satisfaction data
Callcenter(e.g., customer interaction notes)
Inputs from RMs(e.g., sales logs)
Telcos (e.g., top-up patterns, monthly bill payments)
Wholesalers(e.g., paymenthistory for SMEs)
Utilities (e.g., payment record)
Website navigation data
Video analysis of customer footage
Comments on company’s page / website
Social media sentiment
SOURCE: Dave Evans (April 2011) "The Internet of Things: How the Next Evolution of the Internet Is Changing Everything” 9
Where are the opportunities?Three ways advanced analytics is driving value creation
10
Boosting traditional P&L levers• Accelerating growth• Enhancing productivity• Improving risk control Finding new
sources of growth
Delivering the digital bank
Not just FinTechs – the best incumbent credit unions and banks are on the move…
Leading banks are investing
17-20% of their EBIT to support large-scale digital and analytics transformations Redesigning organizations and operating models to
achieve agility and innovation
Building an arsenal of data and analytics capabilities
Radically reshaping key customer journeys
Modernizing their technology infrastructure
11
… focused on driving next generation ofretail banking
SOURCE: McKinsey 12
Call centers
• 50% efficiency gains over next four years
• Powerful source of customer insight for better selling and advice
Bank branches
• 40% efficiency gains over next four years
• Powerful source of customer insight for better selling and advice
Mobile banking
• 70% interactions on mobile over the next four years
Automation
• 70% efficiency gains in the next four years
• Far fewer errors than outsourced workforce
The case for change
Driving analytics led transformation
13
Contents
14
Data gatheringData gathering
Model developmentModel development
Workflow integrationWorkflow integration
Where do most data analytics engagements fail?
Where do most data analytics engagements fail?
Deployment and adoptionDeployment and adoption
Delivering impact requires more than just data and models
“Analytical impact at scale is
• 10% analytics
• 90% end user adoption
Most companies fall short on the latter”
15
Vision and strategy
Organization & Talent
Data
Agile Culture
Modeling tools & techniques
Value assurance
Building a world-class data analytics organization requires focus on 6 key elements
16
Prioritization exercise to assess most impactful analytics use cases and customer journeys and build a roadmap
High
LowLow
Imp
ac
t
Feasibility High
20Prioritized use cases
21
1
26
22719
6
108 2312
3
4 5
7
9
10
11
1314
17
18
19
22
24
25
28 29
Use case prioritization
Feasibility and risk considerations
BusinessWhat economic benefits can be realized (and when)?
Impact considerations
StrategyWhat is the alignment with the business aims and aspirations?
CustomerWhat improvements can be made to customer perception?
DataWhat is the quality of data available and how complex will it be to scale-up data collection and delivery?
CapabilitiesWhere is the organization today relative to aspirations?
Obstacles to changeHow will internal politics/change management perspectives and external environment help or hinder change?
VISION AND STRATEGY
17
Completely decentralised Completely centralisedWith CDO
CDO unit
Data owner units
Several organizational archetypes for analytics
SOURCE: McKinsey Analytics
ORGANIZATION AND TALENT
18
A
B
C
D
Data ScientistData Scientist
Data EngineerData Engineer
Business TranslatorBusiness Translator
What is the most difficult analytics role to hire?
What is the most difficult analytics role to hire?
Workflow IntegratorsWorkflow Integrators
ORGANIZATION AND TALENT
19
Technology skills
Analytics skills
Business skills
DeliveryManagers
B.
Businesstranslator
C.
VisualisationAnalyst
D.
WorkflowIntegrator
E.
Datascientist
F.
Data engineer
G.
Data architect
H.
Business leaders
A.
The most critical talent to find are ‘translators’ who can bridge different functional areas
SOURCE: McKinsey Analytics
ORGANIZATION AND TALENT
20
Customer data
Collateral data
PricingC/C …
BANKING EXAMPLEA strong data foundation is a key enabler
SOURCE: McKinsey Analytics
Governance systems
Legacy systems Customer
dataCollateral
dataPricingC/C …
Data users
Data Warehouse/lake
Risk Accounting …
To parallel modern architectureFrom “spaghetti” architecture
DATA
21
Risk Accounting …
Value leakage in the “last mile” is usually driven by inability to activate the insights appropriately
10
100
Taking Action
Interpreting Insight
Value atStake
Making Decision
OutcomeIncorporating feedback
Monitoring outcome
Value leakage in the “last mile” ($M)
VALUE ASSURANCE
22
An Agile way of working key to bringing multiple skillsets together
SOURCE: McKinsey Analytics
AGILE CULTURE
23
Data Scientist
DataEngineer
DeliveryManager
WorkflowIntegrator
VisualisationAnalyst
Business Translator
DataArchitect
The Data-Driven Credit Union: Powering Transformation with
Advanced Analytics April 20, 2017