Data to Knowledge to Results Review and Analysis of Paper by Davenport et al Team: Something Different Myron Burr Kevin McComas Easwar Srinivasan Bill.

Data to Knowledge to Results Review and Analysis of Paper by

Davenport et al

Team: Something Different

Myron Burr

Kevin McComas

Easwar Srinivasan

Bill Winett

Data vs Information

Data : Measures, Transactions Knowledge / Information

Parts per hour Billing rate Click through rate

Profit maximizing product mix Profit maximizing bundling of solutions Individualized, targeted web pages

What are the Issues?

• Background:– Firms are spending billions on IT applications ( ERP, POS scanners,

web and e-commerce systems, and CRM)– Generated billions of transaction records

• Observation:– Very little data is converted to knowledge (less than 10% in studied

firms)

• Problem Statements:– Lost opportunities for improved results– Unrealized business value from these investments

Proposed Approach to Resolution

Davenport et al, researched over 100 companies

Developed a model for building analytic capability

Demonstrated how to realize results from this capability

Framework

Strategy• What are our core business processes?• What key decisions need analytic

insights?• What information matters?

• Clear strategy leads to good measurements and therefore good data gathering

Context

Process needs a foundation Required ingredients for success Grounded in

Firm’s strategy (and the information needed to execute this strategy)

Skills and experience of staff

Organization and culture Data-oriented / Fact-based Technology and Data

Skills and Experience Key Roles

DB Administrator: loads, organizes and checks data

Business Analyst / Data Modeler Decision Maker / Outcome Manager

Skills: Depth depends on above role Technology Skills Statistical Modeling and Analytic Skills Knowledge of the Data Knowledge of the Business Communication and Partnering

Without skilled staff, IT applications are a waste of $$$.

Organization and Culture

62% of managers: organization and culture biggest barriers to getting significant return on IT investment

Related to skills and experience Value Data-oriented / Fact-based

analysis and decision making Organization of analytics staff

Centralized or decentralized depends on:

Sophistication of the analysis Amount of local knowledge needed Cultural orientation of the firm

Technology and Data

Specific hardware and software, networking and infrastructure

Transaction versus analytic approach

Integration of analytic technologies Requires human insight; can’t

automate 60 to 80% of cost in cleaning up and

integrating data

TransformationData to Knowledge

Analytic and Decision Making Process Depends on experience and relationships

of analysts and decision makers Working closely with decision makers to

understand the questions: Standard, highly-structured: Inventory?

Sales? Semi-structured: Optimum inventory

level? Production versus forecasting? Unstructured: customer segment

migration? An evolving and iterative process Use “decision audits” to evaluate

effectiveness of process

Outcomes

Desired financial outcomes (greater profitability, revenues, or market share) may require changes in: Behaviors: e.g., cost control Processes and Programs: e.g.,

development of new marketing initiative

Extensive communication may be required

Implementation of decisions will determine result.

Application Methodology• Flowchart

High qualitytransaction

data?

Analytical skillsand culture in

place?

Broad need inorganization?

Supportivesenior

executives?

Yes

Yes

Yes

Integrateanalytical

capabilitiesinto business

Implementnew systems

and dataarchitectures

Launch smallpilots andeducate

managers

Launchanalyticalinitiative insingle area

Launchanalytical

organizationalchangeprogram

No

No

No

No

Yes

Implementation Options

• Business needs to dictate extent of implementation and level of focus

Examples

Source: http://www.cs.csi.cuny.edu/~imberman/DataMining/KDD%20beginnings.pdf

More Results

• Earthgrains eliminated 20% of products, increased profits by 70%

• Owens & Minor won $100M contract by showing customer how to save money

• Wachovia Bank improved performance by modeling branch locations

• Harrah’s Entertainment plans to use customer data to increase cross-selling

• Fleet Bank saved >$12M encouraging customers to change from branches to ATMs

Outcome: Increased Profitability

Cumulative Profitability Dependence on Route Complexity

0

5

10

15

20

25

30

35

40

45

50

0 5 10 15 20 25 30 35 40

Number of Routes

Cum

. Pro

fit ($

Mill

ions

)

Other Applications of Data to Knowledge to Results

Source: http://www.cs.csi.cuny.edu/~imberman/DataMining/KDD%20beginnings.pdf

Take-Aways

To get the most from your IT investment:• Hardware, software, networking and

infrastructure only the starting point• You need to commit significant skilled

human resources• Develop sophisticated analytic processes• Instill culture that values data and creating

information• Make decisions on info and then execute

Additional Resources

• SAP.com • Oracle.com • Google Analytics• Accenture.com• Spotfire.com• i2.com• Salesforce.com• cio.com• b-eye-network.com• juiceanalytics.com• WonderWare.com

Questions?