master class davenport

Executive Masterclass

Analytics at Work: Smarter Decisions, Better Results

Thomas H. DavenportPresident's Distinguished Professor of IT and Management

Babson College

Thomas H. Davenport – Analytics at Work

Analytics at WorkAnalytics at WorkyySmarter Decisions, Better ResultsSmarter Decisions, Better Results

Tom DavenportTom DavenportB b C llB b C llBabson CollegeBabson College

SAS PBLS Hong Kong SAS PBLS Hong Kong MasterclassMasterclassSAS PBLS Hong Kong SAS PBLS Hong Kong MasterclassMasterclass12 August 201012 August 2010

From Where Do These Ideas Come?From Where Do These Ideas Come?

• Competing on Analytics: The New Science of WinningWinning• Based on a Harvard Business Review article

in 2006, and an initial study of 32 companies• Strong focus on companies that had made

analytics a key competitive advantage• Led to study of many more companiesLed to study of many more companies,

several surveys, and several industry-specific analyses

Analytics at Work: Smarter Decisions Better • Analytics at Work: Smarter Decisions, Better Results• Addresses how any company can become y p y

more analytical and fact-based• Orientation to the linkage between analytics

and decisions

Thomas H. Davenport – Analytics at Work3 | 2010 © All Rights Reserved.

and decisions

The Decisions Dishonor Roll

►Private sector► Subprime real estate decisions at Lehman Brothers,

Countrywide, Wachovia, Goldman, etc.► The decision to expand in farm equipment at► The decision to expand in farm equipment at

Tenneco► The decision not to sell Yahoo to Microsoft► A certain oil exploration decision by BP

►Public sector► The decision to invade Iraq► The decision to stay in Vietnam and escalate the war► Th d i i t i d C b t th B f Pi► The decision to invade Cuba at the Bay of Pigs► The decisions to launch Challenger, and not to

rescue Columbia


What Do Bad Decision-Makers Have in Common?Have in Common?

►…to use data and analytics

F ily

►…to examine decision alternatives► to have clear decision rolesFailur ►…to have clear decision roles►…to acknowledge human irrationality► t th d i i d

Failure! ►…to agree on the decision made

►…to execute on the decisione!Thomas H. Davenport – Analytics at Work5 | 2010 © All Rights Reserved.

New Decision Frontiers—Are You Exploring?

►Analytics and algorithms► Intuition and the subconscious► Intuition and the subconscious► “The wisdom of crowds”►Behavioral economics and “nudges”►Behavioral economics and nudges►Neurobiology►Decision automation►…Etc.—which are you using?


Deciding on Analytics vs. “the Gut”Deciding on Analytics vs. “the Gut”

40% of major business decisions are not based not on data and facts, but on “ t i ti t”

Statistical predictions consistently outperform

“gut instinct”– Accenture survey

Statistical predictions consistently outperform “gut based” predictions

Extensive evidence that having experts is Extensive evidence that having experts is good, but experts using analytics is much better

Expert intuition is best only when there is little time, limited data and few variables.

“The unexamined life isn’t worth living” S X X

Vdecision making

V

Thomas H. Davenport – Analytics at Work7 7

-- SocratesX X

DecisionDecision--Making in Your OrganizationMaking in Your Organization

Do we spend enough time and attention on key decisions?Do we make decisions in a timely fashion?Do we have clear decision roles most of the time?

Process

Do we generate a variety of decision alternatives?

Are our important decisions made using analysis and data?I t Do we usually have good information to support our decision-

making?Inputs

Are our decision outcomes usually positive?Outcomes


y p

Analytics at WorkAnalytics at Work——The Big PictureThe Big Picture

Analytical Capability Organizational Context Desired Result

A l ti l C lt

DataEnterprise

Analytical CultureAnd Business

Processes

pLeadershipT t

BetterDecisions!

ProcessesTargetsAnalysts .

Systematic Review


What Are Analytics?What Are Analytics?yy

AnalyticsAnalyticsWhat’s the best that can happen?

What will happen next?ge

Optimization

Predictive Modeling What will happen next?

What if these trends continue?

What are the causes and effects?dvan

tag Predictive Modeling

Forecasting

Statistical models What are the causes and effects?

What actions are needed now?tive

Ad Statistical models

Alerts

Where exactly is the problem?

What information really matters?

ompe

tit Query/drill down

Scorecards

What happened?Co

Degree of Intelligence

ReportingStandard reports


Degree of Intelligence

The Planets Are Aligned for AnalyticsThe Planets Are Aligned for Analytics

• ITIT• Data• Skills• Business need• The evidence: “Business

intelligence” was the top spending intelligence was the top spending priority for CIOs in Gartner’s ‘06, ‘07, ’08, and ’09 global surveys07, 08, and 09 global surveys


Levels of Analytical Capability—What’s Yours—Now and Future?What s Yours Now and Future?

Stage 5Analytical

Competitors

Stage 4gAnalytical Companies

Stage 3Stage 3Analytical Aspirations

Stage 2Stage 2Localized Analytics

Stage 1

Thomas H. Davenport – Analytics at Work12

gAnalytically Impaired

Analytical Competitors Analytical Competitors Cut Across IndustriesCut Across IndustriesCut Across IndustriesCut Across Industries

Consumer Products• Procter & Gamble

Telecom• Nextel (not Sprint)• Procter & Gamble

• Mars• Unilever

• Nextel (not Sprint)• Hutchison• CSL• Unilever

Financial Services• Toronto Dominion

• CSLRetail

• J C Penney• Toronto Dominion• BGI/ BlackRock• Progressive

• J.C. Penney• Hudson’s Bay• Kingfisher Asia• Progressive

Government• New York Police Dept

• Kingfisher AsiaTransport / Travel and Entertainment• New York Police Dept.

• VA Hospitals• Hong Kong Efficiency Unit

• FedEx• Hilton


• Hong Kong Efficiency Unit• Octopus Cards

Analytical Companies Perform BetterAnalytical Companies Perform Better

15% of top performers versus 3% of low performers indicated th t l ti l biliti k l t f th i t tthat analytical capabilities are a key element of their strategy.

37%33%

27%

37%

12%8%

19%

9% 10%

0%

8% 9%

No analytical capability

Minimal analytical capability

Some analytical capability

Above average analytical capability

Analytic capability is a key element of

strategy


Source: Accenture Survey of 205/392 companies

The Analytical DELTAThe Analytical DELTA

Data . . . . . . . . breadth, integration, qualityEnterprise . . . . . . . .approach to managing analyticsp pp g g yLeadership . . . . . . . . . . . . passion and commitmentT t fi t d th b dTargets . . . . . . . . . . . first deep, then broadAnalysts . . . . . professionals and amateurs


DataData

The prerequisite for everything analyticalClean, common, integrated Accessible in a warehouseAccessible in a warehouseMeasuring something new and important


Industries and Their Use of DataIndustries and Their Use of Data

Extensive Underachieving Analytical Competitors

Data Streamsfrom Operations

Competitors

FSRetailp

and CustomerRelationships CPGHealth Care

Limited

Disadvantaged Overachieving

Limited Extensive


Use of Data for Analysis and Decision-Making

What Gets in the Way of Great Data?What Gets in the Way of Great Data?

Not owning the data for key functions/relationshipsPharma, grocery/CPG, autos

Lots of M&A activityTelecom, large banks

Rapid change in operational or delivery technologiesg y gTelecom, newspapers, retail

Lack of data standards across the industryLack of data standards across the industryTelecom, health care, books


Data Through the StagesData Through the Stages

Stage 3Analytical Aspirations

Stage 4Analytical

Companies

Stage 5Analytical

Competitors


Stage 4Analytical

Companies

Stage 5Analytical

Competitors


Stage 4Analytical

Companies

Stage 5Analytical

Competitors


Stage 4Analytical

Companies

Stage 5Analytical

Competitors


Stage 4Analytical

Companies

Stage 5Analytical

Competitors

Stage 1Analytically Impaired

Stage 2Localized Analytics

Analytical Aspirations













Stage 4 Stage 5Stage 3 Stage 4Stage 2 Stage 3Stage 1 Stage 2

Analytically Impaired to Localized Analytics

Localized Analytics to Analytical Aspirations

Analytical Aspirations to Analytical Companies

Analytical Companies to Analytical CompetitorsLocalized Analytics

•Gain mastery over local data of importance,

including building functional data marts


• Build enterprise consensus around some analytical targets and their data

needs

Analytical Companies

•Build enterprise data warehouses and integrate

external data.

Analytical Competitors

•Educate and engage senior executives in competitive

potential of analytical data.data marts. needs.

• Build some domain data warehouses (e.g.,

customer) and

•Engage senior executives in EDW plans and

management.

•Exploit unique data.

•Establish strong data governance, especially

corresponding analytical expertise.

• Motivate and reward cross-functional data

•Monitor emerging data sources.

stewardship.

•Form a BICC if you don’t have one yet.


functional data contributions and

management.

EnterpriseEnterprise

Enterprise perspectives on:D tDataAnalystsTechnology

Which do you have?c do you a e


E Is also for “Expense” in 2010E Is also for “Expense” in 2010

Consolidate reporting and analytical software across the enterprisesoftware across the enterpriseUse software you already haveApply analytics to particular decisions, so you can measure and justify the expenseyou can measure and justify the expense


Role of LeadershipRole of Leadership

What great analytical leaders do…

Demonstrate persistence over time Push for more data and analysisWork along multiple fronts Build an analytical ecosystemBuild an analytical ecosystemSet strategy & performance expectations Hire smart people & give them credit for being smartLook for leverageSet a hands-on exampleSi f ltSign up for resultsKnow the limits of analyticsDevelop their people skills


Develop their people skillsTeach

22

Analytical LeadersAnalytical Leaders

Shannon Antorcha of Carnival Cruise Lines – Analytical Department Leader “If you’re going to be a change agent, you have to educate people and help them understand what

you’re trying to do. Eventually you will get their buy-in.”

Greg Poole of The Talbots – Business Function LeaderGreg pushes for more data and analysis by communicating key metrics and posts charts and graphs in G eg pus es o o e data a d a a ys s by co u cat g ey et cs a d posts c a ts a d g ap s

his office.

Tom Anderson – Division Head and Entrepreneur“The beauty of analytics, is that you find lots of things that can be incrementally improved”

Jim and Chris McCann – CEO and President of 1800Flowers“W h lt f l ti d t ti I ‘I k h t thi k t ll h t


“We have a culture of analytics and testing. I say – ‘I know what you think – tell me what you can prove.’”

23

LeadersLeaders Set Set anan Example Example


TargetsTargets

• Support a key strategic capability

• Engage top management commitment

• Create momentum for analytics across the yenterprise

• Have ambitious (business impact) yet pragmatic scope

• Are data rich – or have the potential to beThe Best Targets…The Best Targets…• Dramatically improve effectiveness of asset and/or

labor-intensive activities

• Have broad implications across functions, processes, geographies or business units.


TargetsTargets——How High Are Your Sights?How High Are Your Sights?

Optimal response

embedded in real-time I tit ti l A ti

Real-Time Optimization

real time process Prediction and

differentiated action

embedded in process

Institutional Action

Predictive ActionPredictions of response by

target/ segment

process

Different h f

Differentiated Action

Key targets and segments

approaches for different targets/

segmentsKey Targets/Segments

segments defined

Well-defined, common, clean, and integrated

Data in Order


and integrated data

Targets Spread Across IndustriesTargets Spread Across Industries

Yield management/price optimizationYield management/price optimizationAirlines Hotels Retail Insurance

Randomized testing with controlsAgriculture Pharma Retail Onlineg

Behavioral targetingDi t il O li d S i l diDirect mail Online ads Social media

What are other industries doing today that you’ll do tomorrow?


you ll do tomorrow?

AnalystsAnalysts

1%Analytical Champions--OwnL d l ti l i iti ti

5 10%Analytical Professionals—Own/RentC t l ith

1% Lead analytical initiatives

5-10% Can create new algorithms

Analytical Semi-Professionals—Own/RentyCan use visual and basic statistical tools, create simple models

15-20%

Analytical Amateurs--OwnCan use spreadsheets, use 70 80% Can use spreadsheets, use analytical transactions70-80%


* percentages will vary based upon industry and strategy

Typical Skill Level by Type of AnalystTypical Skill Level by Type of Analyst

Quantitative Business Relationship and Coaching and knowledge and

designconsulting staff development

Amateur

Semi-professional

Professional

Champion


ExpertIntermediateBasic Foundational Advanced

Five Ways to Organize AnalystsFive Ways to Organize Analysts


Analyst Organization and EngagementAnalyst Organization and Engagement


Analyst Organization and PersistenceAnalyst Organization and Persistence


DELTA Stage ModelSuccessFactor




Stage 4Analytical Companies

Stage 5Analytical Competitors

Data Inconsistent, poor quality and organization; difficult to do substantial analysis; no groups with strong data orientation.

Much data useable, but in functional or process silos; senior executives don’t discuss data management.

Identifying key data domains and creating central data repositories.

Integrated, accurate, common data in central warehouse; data still mainly an IT matter; little unique data.

Relentless search for new data and metrics; organization separate from IT oversees information; data viewed as strategic assetas strategic asset.

Enterprise No enterprise perspective on data or analytics.Poorly integrated systems.

Islands of data, technology, and expertise deliver local value.

Process or business unit focus for analytics. Infrastructure for analytics beginning to coalesce

Key data, technology and analysts are managed from an enterprise perspective.

Key analytical resources focused on enterprise priorities and differentiation.

coalesce.

Leadership Little awareness of or interest in analytics.

Local leaders emerge, but have little connection.

Senior leaders recognizing importance of analytics and developing analytical

Senior leaders developing analytical plans and building analytical capabilities.

Strong leaders behaving analytically and showing passion for analytical competition.

capabilities.

Targets No targeting of opportunities.

Multiple disconnected targets, typically not of strategic importance.

Analytical efforts coalescing behind a small set of important targets.

Analytics centered on a few key business domains with explicit and ambitious outcomes.

Analytics integral to the company’s distinctive capability and strategy.

Analysts Few skills, and those attached to specific functions.

Unconnected pockets of analysts; unmanaged mix of skills.

Analysts recognized as key talent and focused on important business areas.

Highly capable analysts explicitly recruited, developed, deployed, and engaged.

World-class professional analysts; cultivation of analytical amateurs across the enterprise.


Your Organization’s DELTAYour Organization’s DELTA

What are your organization’s DELTA strengths?On which DELTA factors could you stand some improvement?Where are you currently focusing your efforts?Can you share some of your approaches to addressing a DELTA factor or two?


The Context: Analytical CultureThe Context: Analytical Culture

Facts, evidence, analysis as the primary , , y p yway of decidingPervasive “test and learn” emphasis where th ’t f tthere aren’t factsFree pass for pushbacks—”Where’s your data?”data?Still room for intuition based on experienceA focus on action after analysisA focus on action after analysisNever resting on your analytical laurels


The Context: Analytical ProcessesThe Context: Analytical Processes

Inventory ForecastInventory Forecast

Defection RiskDefection Risk“What is the customer status?”

Gl b l ATPR tCreation

CreationPurchase Order

Inventory ForecastInventory Forecast“Will this be back in inventory?”

Global ATPCheck

RequestGlobal ATP

CreationSales Order

Fulfillment Request

Creation &Release Delivery

DeliveryExecution

CLTVCLTV“Does this order justify extra

efforts?”

yRequestReturns per CustomerReturns per Customer

“What is the customer history?”

Releases ASNUpdate

Inventory Accounting

UpdateInventory

Receives ASNDelivery PerformanceDelivery Performance

“How effective is our fulfillment process?”


Source: SAP AG 2006

A Study of DecisionsA Study of Decisionsyy

►57 attempts to improve specific decisionsp p p►90% of companies could name one►Most decisions were frequent andDecisions! ►Most decisions were frequent and

operational► Pricing (of consumer goods, industrial goods, government g ( g , g , g

contracts, maintenance contracts, etc.);► Targeting of consumers for marketing initiatives (by retailers,

insurers, credit card firms);insurers, credit card firms);► Merchandising decisions by retailers (what brands to buy in

what quantity for what stores, shelf space allocation);► L ti d i i (f b k b h h t i ► Location decisions (for bank branches, where to service

industrial equipment)

►Results in “Make Better Decisions,” Harvard


►Results in Make Better Decisions, Harvard Business Review, Nov. 2010

Systematically Making Decisions BetterSystematically Making Decisions Better

IdentifyIdentify InventoryInventory

Better Decisions

Better Decisions

InterveneIntervene InstitutionalizeInstitutionalize


Identify Your Most Important DecisionsIdentify Your Most Important Decisions

Whi h 5 t 10 d i i t iti l t Which 5 to 10 decisions are most critical to your strategy?

1What businesses should we be in?What is the customer value proposition? 123How do we gain share and grow?

Which two or three operational decisions are critical to pthe execution of each strategy decision?

What price should we charge?p gWhich new product should we move forward?

Can you share any strategic or tactical decisions with


Can you share any strategic or tactical decisions with us?

2010 © Thomas Davenport. All Rights Reserved.

Identifying Key Decisions at a Pharma FirmIdentifying Key Decisions at a Pharma Firm

St t iStrategicWhat product categories?

Led to newWhat therapeutic areas?What geographical markets?

Led to newstrategy process

What mix of project activities?

OperationalpWhich compounds to move through phases Led to full

lifec cle Which indications to pursueHow to position drugs relative to

tit

lifecycle responsibility


competitors2010 © Thomas Davenport. All Rights Reserved.

Inventory Key DecisionsInventory Key Decisions

Who’s responsible for it (“Who Who’s responsible for it (“Who Has the D?”)H ft i it d ?How often is it made?How long does it take?What process is being used?How well does it work?How well does it work?Does it need an intervention?

H h i iti d d i i ?Thomas H. Davenport – Analytics at Work

2010 © Thomas Davenport. All Rights Reserved.

How have you prioritized your decisions?

Inventorying New Product Development Inventorying New Product Development Decisions at ETSDecisions at ETSDecisions at ETSDecisions at ETS

More competition for key test franchises made new More competition for key test franchises made new product development decisions particularly important“L ” f d t i th t“Long runway” for new products in the past“Stage gate” process, but matrix structure led to

l ibiliti f ti d i iunclear responsibilities for gating decisionsLack of clear information about IP, potential partners, and likely markets, so created “rubric”Team examined process, and leaders of team took ongoing responsibility for decisionsGreatest success so far: killing truly bad ideas

Thomas H. Davenport – Analytics at Work2010 © Thomas Davenport. All Rights Reserved.

quickly

Types of InterventionsTypes of Interventions

New analytical techniques (telecom equipment firm)New metrics and data (e.g., optical firm)New data repositories (vaccine firm)New data repositories (vaccine firm)New decision systems (P&C insurance information supplier)Knowledge-sharing approaches (tools firm)Knowledge sharing approaches (tools firm)Change in the business process involved in the decision (retailer)Education of decision-makers and related function (insurance firm)Education of decision-makers and related function (insurance firm)Communications initiatives about the decision (ad agency)New methods and theories (brokerage firm)New methods and theories (brokerage firm)A culture of honesty and decisiveness (oil company)Which have you employed?


Which have you employed?2010 © Thomas Davenport. All Rights Reserved.

Most Common Decision InterventionsMost Common Decision Interventions

0,9

0,7

0,8

0,5

0,6

y Men

tioni

ng

A b 0,3

0,4

Freq

uenc

yAverage number mentioned

per decision:

0,1

0,2per decision:

5.3!

0


Type of Intervention

Linking Information and DecisionsLinking Information and Decisions


Technologies Linking Technologies Linking Information and DecisionsInformation and DecisionsInformation and DecisionsInformation and Decisions


Your Organization’s LinkageYour Organization’s Linkage

How do you link information and decisions?decisions?Have you tried to create closer linkages?What technologies have you employed?


Multiple Interventions: Multiple Interventions: Better Pricing Decisions at StanleyBetter Pricing Decisions at StanleyBetter Pricing Decisions at StanleyBetter Pricing Decisions at Stanley

Pricing identified as one of four key decision domainsPricing identified as one of four key decision domainsPricing Center of Excellence established in 2003Adopted several difference pricing methodologiesImplemented new pricing optimization softwarep p g pRegular “Gross Margin Calls” for senior managersOffshore capability gathers competitive pricing dataOffshore capability gathers competitive pricing dataSome automated pricing systems, e.g., for

tipromotionsCenter spreads innovations across Stanley


Result: gross margin from 34% to over 40% in six years

Institutionalizing Better DecisionsInstitutionalizing Better Decisions

Decision Decision coaches/consultants/analysts to helphelpDecision improvement methodologiesmethodologiesEducation and guidelines for managersmanagersPost-decision reviewsWhich have you employed?What’s worked, what hasn’t?


Key Decision Analyst/Coach RolesKey Decision Analyst/Coach Rolesy yy y

“Help to frame the decision”

“Stand firm when necessary”

“Tell a story with data”

“Build a rapid prototype”


Build a rapid prototype

Institutionalizing Better Decisions at ChevronInstitutionalizing Better Decisions at Chevrongg

Decision analysis group gets attention by Decision analysis group gets attention by recommending against refinery projectDA group begins to lead decision workshopsDA group begins to lead decision workshopsBuilds and refine economic and analytical modelsAll j t $100M i d i i l iAll projects over $100M require decision analysisEx post facto assessment of decision quality

i d f l j trequired for large projectsDA group has trained more than 2500 decision-makers, and has certified 10,000 (including the CEO) through online training module


Culture of honesty and self-examination

Barriers to Better DecisionsBarriers to Better Decisions

There are too many decisions to addressPeople don’t want other people intervening in their mental processesPeople don t want other people intervening in their mental processesMost decision-makers would rather avoid accountabilitySenior executives will feel that this is their territoryBecoming too engineering-focused might limit creativity


Decision technologies not well-developed

Roles for IT in Improving DecisionsRoles for IT in Improving Decisionsp gp g

Restructure the entire IT organization to emphasize decision-making

e.g., P&G’s “Information and Decision Solutions”

Establish a COE, competency center, or consulting group around analysis and decisions

e.g, Kimberly-Clark’s BICC

Include analytics and decision processes in the y pbroader information provision process

E.g., Cisco Advanced Services “Production Analytics”

At the very least, ask “What decision does this support?” when asked to provide information


Next Steps for Next Steps for Analytical DecisionsAnalytical Decisions

Continual pursuit of new pmeasures and data types

RFID and sensorsVoice, video, text

Further integration with Further integration with decision automation and decision managementdecision managementKnowledge

t/ l ti l management/analytical resource management


Social analytics

It Doesn’t Happen Overnight It Doesn’t Happen Overnight —— Start Now!Start Now!

Takes a while to put data and infrastructure foundation in place, and even longer to develop human capabilities, a fact-based culture, and “success stories”Barclay’s five-year plan for “Information-Based Customer Management”UPS — “We’ve been collecting data for six or seven years, but it’s only become usable in the last two or th ith h ti d i t lid t three, with enough time and experience to validate conclusions based on data.”


master class davenport

Documents