Accenture counting on_analytical_talent

Counting on Analytical TalentBy Jeanne G. Harris, Elizabeth Craig and Henry Egan

March 2010

Contents

2 | Accenture Institute for High Performance | Copyright © 2010 Accenture. All rights reserved.

Executive Summary ............................................................................................................... 3

How to Create a Talent-Powered Analytical Organization .......................................... 7

How to Organize Your Analytical Talent .......................................................................... 17

How to Engage and Retain Your Analytical Talent ....................................................... 27

About the Research ............................................................................................................... 40

About the Authors ................................................................................................................. 46

Notes ......................................................................................................................................... 47


Executive Summary

Counting on Analytical Talent

Companies are increasingly turning to analytics to gain acompetitive edge. As they do, they must resolve unique demandson their information technology, their structure, their processes,and their culture. Most critical, however, is the challenge posedby analytical talent, the people at all levels who help turn datainto better decisions and better business results.

In this report, we explain how companies are harnessing the“talent power” of their analysts, organizing them most effectively,and keeping them engaged for the long term. For companiesthat want to succeed with analytics, it’s essential that theyrecognize analytical talent as a distinct workforce, and managethem in accordance with their growing importance.

Are your analysts working on yourmost important problems? Are theyturning operational data into insightsyou can act on to outperform yourcompetitors? Can you count on themto help you make better decisions?

The answers may lie in how well youare managing your “analytical talent.”These are the people who use statistics,rigorous quantitative or qualitativeanalysis and information-modelingtechniques to shape and make businessdecisions—the “quant jocks,” “mathbrainiacs,” “Excel ninjas” and otheranalysts who bring you the data, thequantitative analysis and the statisticalmodels you need to improve decisionsabout everything from new productofferings to marketing investments.

And more and more, such people matterto performance because analytics itself,as a way of making smarter businessdecisions and getting better results,matters more and more. In a recentAccenture survey of executives at largecompanies in the US and UK, nearlythree-quarters of participants saidthey are working to increase theircompany’s use of analytics.1

For these reasons, more than two yearsago, we began studying analytical talentas a distinct group. One of us, Jeanne,was already a worldwide expert onanalytics, having just published agroundbreaking book, Competing onAnalytics, on how organizations use data-driven insights to generate impressivebusiness results.2 And she was already

hard at work on the follow-up,Analytics at Work, which reveals howorganizations can effectively deployanalytics in their day-to-day operations—one business decision at a time.3

Elizabeth, meanwhile, had studied talentfor years, as a professor and then asan Institute researcher and co-authorof The Talent Powered Organization,which explains how companies cantransform talent management from asupporting function to an essentialcompetitive capability.4 Henry, ourthird author, was in fact our ownanalytical talent, adept at makingsense of large quantities of data andderiving insights from the results.

Having assembled this team, we startedwith some basic questions about theseincreasingly important employees:Who are they? What motivates them?What makes them effective? Howwell do companies manage them? Weinterviewed dozens of executives inanalytical organizations and surveyedmore than 1,300 employees to investi-gate these questions.

What we found was surprising.

First, we learned that companies neglectthis group. They don’t see analyticaltalent as a distinct and valuableworkforce—and they certainly don’tmanage it as such. Analysts are oftenscattered throughout departments;many companies don’t have a clearpicture of who their analysts are orwhere they reside organizationally.In fact, most companies have manydifferent job descriptions for similaranalyst roles—if they have them at all.As a result, performance expectationsand measurements are vague or



inconsistently applied. And analysts’daily work activities are not alwaysaligned with the strategic goals ofthe organization. This is a recipe fordisaster when it comes to attracting,engaging and retaining analyticaltalent and building an organization’sanalytical capability.

How can companies overcome thisneglect of analytical talent and get themost out of this critical workforce?As we discuss in part one of this report,the key is to take an enterprise-wideapproach to managing analysts. Wediscovered that the most successfulcompanies create a talent-poweredanalytical organization by building fourtalent management capabilities:defining analytical talent needs;discovering new sources of analyticaltalent; developing analytical talent;and deploying analytical talent effec-tively. In doing so, they not only addressthe strategic and operational needs ofthe business but also unleash theiranalysts’ talents to continually expandthe company’s analytical capabilities.

Second, we learned that companiesoften struggle with how to organizeanalysts. Should they be centralized ordecentralized? “Charged out” to therest of the business as consultantsor made available as a free resource?Where and to whom they shouldreport? Most companies use one offive models to organize their analysts,often depending on the organization’s

relative degree of analytical sophisti-cation. But their choices are oftenproblematic. When companies don’tget it right, their best analytical mindsare relegated to conducting simpleanalyses or working on low-valueprojects rather than building robustmodels to solve the most challengingbusiness problems. Worse still, thattype of work is a sure recipe for analystdisengagement and defection.

As we explain in part two of this report,our research revealed that companiesthat want to build a strong analyticalworkforce are best served by greatercentralization and coordination oftheir analytical talent. Doing so ensuresthat analysts are working “close to thebusiness” on the most important initia-tives and also “close to one another” tocoordinate their efforts and to promotemutual learning and support. It alsoensures that analysts have the kind ofmeaningful work and career opportunitiesthat are critical to their engagementand retention.

The third surprising discovery of ourresearch was that while analysts aremotivated by many of the same thingsas other employees, some things areuniquely important to analytical talent.Three things are essential for engagingand retaining all employees: Companiesneed to provide meaningful work andcareer opportunities, support people’sefforts as well as their need forrecovery and renewal, and cultivate aculture of trust and respect. But analystshave a different view of what makesopportunities meaningful, what kindsof support are essential, and what

cultural factors matter most. Analystsare most engaged when they under-stand the business side of things aswell as the analytics, when they knowwhat is expected of them, and whenthey can keep their technical skills andexpertise current. They are most likelyto stay when they have a high degreeof management support. We examinewhat engages analysts, and what makesthem likely to stay with an organization,in part three of this report.

As companies continue to seek com-petitive advantage and ways ofdifferentiating themselves, they willbe counting on analytical talent morethan ever. Your company’s successwith analytics hinges on its ability toeffectively manage analysts, so youneed to understand what makes themtick. This report won’t tell you every-thing you need to know, but it will getyou started. Regardless of whether yourcompany routinely uses analytics as adistinctive business capability or is justbeginning to develop analytical aspira-tions, the care and feeding of analyticaltalent is critical to your success.





How to Create a Talent-PoweredAnalytical Organization



Successful companies have learned how to harness the powerof this increasingly important kind of talent. To get started,you have to recognize the “quant jocks,” “math brainiacs,”“Excel ninjas” and others who regularly work with data as adistinct workforce—one that is pivotal to an organization’ssuccess. Our research reveals that generating talent powerwith this workforce comes from building four key talentmanagement capabilities.

Few companies manage analyticaltalent as a strategic resource. Becauseanalysts are often scattered throughoutthe organization, many companiesdon’t even have a clear picture of whotheir analysts are, where they resideorganizationally or exactly how manythey have. They certainly don’t recognizeor manage them as a distinct and criticalworkforce segment that requires itsown recruiting strategies, training anddevelopment plans, career paths orperformance management processes.In fact, more often than not, companieshave many different (or poorly written)job descriptions for similar analystroles—if they have descriptions at all.Performance expectations and mea-surements are vague or inconsistentlyapplied. And analysts’ daily work activitiesare not always properly aligned withthe strategic goals of the organization.

This is not just poor HR practice; it’sa recipe for disaster when it comesto attracting, engaging and retaininganalytical talent and building anorganizations’ analytical capability—a goal that many large companiesaspire to meet in the near future.5

Our research sheds light on how toovercome this neglect of analyticaltalent and to get the most out of thisincreasingly critical workforce (Seethe Appendix, “About the Research”).Through dozens of interviews withsenior executives in successful compa-nies and an extensive survey of analystsfrom a wide range of industries, wediscovered that the most successful

companies create a talent-poweredanalytical organization by building fourkey talent management capabilities:

1. Defining analytical talent needs:What types of analysts do you need?

2. Discovering new sources of analyticaltalent: Where do you find top analysts?

3. Developing analytical talent: Whatskills do analysts need and how doyou build them?

4. Deploying analytical talent: Howdo you create the best possible matchbetween analysts’ skills and businessdemands?

What distinguishes talent-poweredanalytical organizations isn’t just thequality of their analytical talent: it’s



their ability to unleash their analysts’talents to maximize and continuallyexpand the company’s analytical capa-bilities. Here’s what you need to knowto create your own talent-poweredanalytical organization.

Define your analyticaltalent needs

This step requires a deeper under-standing of the variety of analyticalroles and the types of analytical talent.In a large corporation, hundreds ofemployees have the word “analyst”in their functional titles.

We define “analysts” as workers whouse statistics, rigorous quantitativeor qualitative analysis and information-modeling techniques to shape and

Analytical ChampionsLead analytical initiatives

1%

5-10%

15-20%

70-80%

Analytical ProfessionalsBuild analytical models and algorithms

Analytical Semi-ProfessionalsApply analytical models to business problems

Analytical AmateursPut the output of analytical models to work

Types of analysts

Percentages represent the proportions of different types of analytical talentin a typical organization.



make business decisions—still a broadrange of activity. In our research andexperience with dozens of analyticallyoriented companies in many industries,we’ve determined that successfulanalytical organizations need fourtypes of analytical people: champions,professionals, semi-professionalsand amateurs.6

ChampionsChampions are senior executiveswho rely on rigorous data and analysisto run their business units. They areresponsible for aligning analytics withbusiness strategy. In some organizations,like insurance companies and riskmanagement firms, executive titles suchas “senior vice president of customerinsight and analytics” or even “chiefanalytics officer” are not uncommon.Champions are the leading advocatesand advisers on how analytical tech-niques and technologies—such as trend-mapping and forecasting, predictivemodeling and enterprise-resourceplanning systems—can be used to guidedecision making.

An appreciation of analytics, ratherthan a wealth of technical know-how,is the central requirement for thisrole. As a result, while some championshave deep technical knowledge and theacademic credentials to match, themajority come from a business back-ground. (See “Help Wanted: AnalyticalChampion.”)

ProfessionalsProfessionals possess the deepestquantitative skills. They are the chiefarchitects of analytical applications,developing the statistical modelsand algorithms used by others in theorganization. Professionals typicallyemploy complex techniques—such astrend analysis, classification algorithms,optimization and simulation—and haveadvanced technical skills, includingcoding C+++, SQL and SAS.

These jobs require an advanced degree(often a Ph.D.) in a quantitative field,such as statistics, mathematics, eco-nomics or operations research, or aspecialty degree in a field such asbiostatistics, informatics, genetics orapplied physics. (See “Help Wanted:Analytical Professional.”)

Semi-professionalsSemi-pros are responsible for applyingthe models and algorithms developedby professionals. The majority offinancial and marketing analysts aresemi-pros. They may be sophisticatedanalysts in their own right, and maydevelop applications on occasion, buttheir primary role is to apply analyticsto business problems. They are expertsin data collection, interpretation anduse. Semi-pros are adept at workingwith analytical applications, visual toolsfor information analysis and “what-if”tools, including marketing workbenchesand models for financial planning,pricing or forecasting sales. They arethe primary users of statistical softwarepackages, such as SAS or SPSS, orenterprise systems like SAP.

At the interface between analyticsand the rest of the business, semi-prosneed to be able to translate the resultsof analyses and the benefits of analyticsinto layperson’s language. Some semi-pros are primarily business analysts,such as MBAs with quantitative orien-tations. (See “Help Wanted: AnalyticalSemi-professional.”)

AmateursAmateurs’ primary focus isn’t necessarilyanalytical. Nevertheless, they need tounderstand analytics to do their jobs.Amateurs might include a call-centeremployee who relies on a “next bestoffer” recommendation in order to servea customer effectively, or a regionalmanager for a hotel chain who moni-tors and occasionally overrides thecompany’s revenue management system.

They are typically businesspeoplewho can enter and manipulate datausing Microsoft Excel spreadsheetsand other basic information manage-ment tools and then put the outputof analytical models to work. The title“amateur” is not meant to suggestthat these are junior employees. Infact, they include many of the mostinfluential employees and executivesin your business. (See “Help Wanted:Analytical Amateur.”)



Help Wanted: Analytical Professional

Lead artifical intelligence—data mining engineer

Job Description:• Responsible for research, design, prototyping and support

of intelligent information processing applications• Responsible for both application and research across

a broad range of AI technologies: data mining, naturallanguage process, human centered design, learning andintelligent systems, and information semantics

• Must be able to design, prototype and support intelligentinformation processing applications with primary focus ondata-mining components and integration and evaluationof statistical analysis technologies

Requirements:• PhD strongly preferred, with 5 years of work experience• Strong skills in data mining and the intersection with

other technologies such as signal and image processing—statistical classification models, clustering, time series data,handling huge data sets, etc

• Ability and desire to lead technical project teams• Ability and desire to meet with customers, understand

their needs, and generate appropriate, innovativetechnical solutions

• Programming ability is desirable: C, Matlab, Java, Perl,Unix/Linux

Help Wanted: Analytical Champion

Director or senior vice president of client analytics

Job Description:• Lead the development of a robust analytical environment• Manage all client analytics projects to ensure timely

delivery of deep client insight• Develop knowledge infrastructure to support client

development• Provide analytical and statistical consulting as required• Perform trend analysis to support decision rationale• Direct the design and implementation of Client

Satisfaction Measurement programs

Requirements:• Minimum of 10 years work experience, the majority of

which should be in an analytical role within a largefinancial services organization; experience in managingan analytical team is preferred

• Proven experience in independently developing robuststatistical models

• Superior hands-on knowledge of data-mining and businessintelligence tools, particularly SAS and SQL

Help Wanted: Analytical Amateur

Supervisor, distribution plant

Job Description:• Monitors throughput, volume fluctuations and labor

utilization to ensure smooth distribution operations.Ensures appropriate staffing to meet new businessrequirements and volume fluctuations.

• Operates within established budgets, forecasts,cost objectives and service levels to control expenses.Supports implementation of cost, quality and serviceimprovement initiatives.

• Supports all established plant and distribution goalsand objectives. Apprises manager of team progressand any obstacles.

• Ensures effective and timely communication of company,plant and department information to managers, peersand work team.

Requirements:• Strong leadership, team-building, and problem-solving skills• Ability to coach and mentor employees• Ability to provide technical leadership

Help Wanted: Analytical Semi-professional

Manager, market research

Job Description:• Manage all marketing tests and analysis including

market-based analysis, field, direct mail ande-commerce tests.

• Define test objectives, key metrics, and test parameters;assure that tests are planned and executed effectivelyto maximize confidence in results.

• Interpret results and recommend strategies, segmenta-tions and programs to enhance customer insights andidentify untapped market demands.

Requirements:• Bachelors Degree, preferably in a quantitative field.• 5+ years experience in a testing and measurement

or data analysis role within an in-house consultativeresearch organization or agency.

• Proficiency in SAS or SPSS• Masters degree strongly preferred.



How many analysts of each type youneed will depend on your industry,business strategy and the analyticalorientation of your company. Thechallenge is to manage the types in away that reflects the unique contribu-tions they make to the organization’sperformance. (See Figure 1).

When managing an analytical workforce,remember that the pros and semi-prosare the ones who create the moststrategic value for your company.Champions are important: They providethe leadership, direction and impetusrequired to execute analytical strategies.But they are few and far between.And while amateurs may use data andanalytics to perform their everyday jobs,they do not possess the rare and valuablespecialized skills that are the lifebloodof the analytical organization.

As you start to think through how tomanage analytical talent more strate-gically, you should focus your effortson your ranks of pros and semi-pros.These scarce and highly-specializedworkers are critical to maintain yourorganization’s analytical capabilities,and they are also among the mostchallenging to attract, develop, engageand retain. Because of their value tothe firm, you can’t simply managethem the same way you manage therest of your employees. You need adifferentiated, strategic, enterprise-wideapproach to these specialists.

Let’s take a closer look at how com-panies can manage these groups mosteffectively.



Discover new sources ofanalytical talent

As you start to ramp up your analyticalcapabilities, it’s inevitable that demandfor analytical pros will outstrip supply.In order to manage this imbalance, youcan give certain analytical initiativespriority over others and train front-linestaff to carry out basic analyses. Yetat some point, to boost the supply andincrease the quality of your analysts,you’ll need new strategies for locatingand attracting the necessary skills.

Look inside your organizationIf your company has just started usinganalytics to create business value, youshould first look inside your organiza-tion to identify the analysts you’vealready got. Scan the functions andbusiness units where you’d expect tofind quantitative talent—those thattraditionally use analytics most, suchas finance, IT, sales and customerservice. However, don’t stop searchingthere. While the majority of analyticalpros and semi-pros will be foundhere, analysts are likely to be widelydispersed across the business.

Consider Bank of America, which beganits efforts to better manage analyticaltalent by characterizing a subset ofits job descriptions as “analytical.” Thebank ultimately identified more than2,000 analytical professionals, semi-prosand amateurs. The outcome of such anassessment can help an organizationreorganize its analysts or focus onrecruiting and training to fill analyticalskill gaps.

Figure 1: Typical skill proficiency levels by type of analyst

Few analysts possess the full range of skills needed to plan and execute majoranalytical initiatives. Therefore, companies need the right mix of analytical talent.

Champion

Professional

Quantitative Businessknowledgeand design

Relationshipand consulting

Coachingand staffdevelopment

Semi-professional

Amateur

Basic Foundational Intermediate Advanced Expert

Look beyond your organizationYou can readily find analytical prosand semi-pros where they naturallygather: at analytics conferences likeINFORMS (the operations researchsociety), vendor-sponsored meetingsand alumni organizations. Manyanalysts tend to live near universitieswith strong quantitative programs orin major financial centers, where theirskills are in high demand. Advancedskills can also be found on socialnetworking sites like LinkedIn, specialtysearch firms and websites such asquantfinancejobs.com or quantster.com.Poaching top analysts from high-performing competitors has alsobecome commonplace.

To secure an ongoing supply of prosand semi-pros, you might forge linkswith top graduate schools. Sponsorshipsand internships can help build closerelationships with academic institutions.Dow Chemical, for example, enjoys along-standing partnership with CentralMichigan University and hires many ofits graduates. Similarly, SAS endoweda “Masters in Advanced Analytics”program at North Carolina State.

If you can’t recruit the analytic talentyou need directly, consider outsourcingsome of your analytic work. An opera-tions executive from a U.S. retailerexplains why that approach suits hiscompany: “The forecasting groupinvolves very technical kinds of skillsand experience, and so I prefer havinga resource that can provide me withthe very best people on an ongoingbasis and has a bigger pool of talentto draw from."7

In the past, companies usually out-sourced analytical work to specialistfirms in North America or WesternEurope. But given the talent shortagein these markets, businesses areincreasingly looking for help in emergingmarkets, such as China and India.Often dismissed as vast pools of cheapnumber-crunchers, these markets arebecoming excellent sources of talent asthey gain the capabilities and experienceto carry out some of the most complexanalytical tasks. Wachovia Bank enlistedGenpact, India's largest business processoutsourcing firm, to provide investmentbanking analysis and other analyticsfor the bank. By 2011, we estimate thatthe majority of the offshore market foranalytics delivery will be based in India.

Some of your most talented analyticalprofessionals don’t even have to beon your payroll. Companies like P&G,Amazon, Eli Lilly and Solvay arefinding creative ways to harness ana-lytical expertise through virtual collab-oration and crowd-sourcing techniques.For example, they are using Web-based “idea marketplaces,” such aswww.Innocentive.com andwww.NineSigma.com to post requests(and rewards) for solutions to theirtrickiest analytical problems. And Netflixheld a competition that offered a topprize of US$1 million to anyone thatcould improve—by at least 10 per-cent—the accuracy of Cinematch, itsmovie recommendation algorithm.8



Develop your analyticaltalent potential

Analytical work requires specializedskills, and the skill requirementschange rapidly as new analytical toolsand techniques emerge. You mustkeep your analysts’ technical skills upto date. But the ability to tune aregression equation or manipulate aspreadsheet is only the beginning.Effective analysts are proficient notonly with data but also with people.Therefore you must invest in developingthe softer skills analysts need tosucceed, as well.

Quantitative and technical skillsThese are the foundation. Naturally,analytical professionals have morequantitative expertise than semi-pros,champions, and amateurs, but allanalytical people must be proficientin the numerical disciplines specificto their industry or business function:stochastic volatility analysis in finance,biometrics in pharmaceuticals andinformatics in healthcare firms, forexample. Analysts must also know howto use the software tools associated withtheir type of analytical work, whetherit be to build models, define decision-making rules, conduct “what if” analysesor interpret a business dashboard.

One obvious approach to developingquantitative and technical skills is toprovide specialized training. At Procter& Gamble, for example, the centralProduct Supply Analytics group offersa course, “Analytics with Spreadsheets.”The trucking firm Schneider National’scentral analytics group offers coursessuch as “Introduction to Data Analysis”and “Statistical Process Controlin Services.”

Developing your analysts has manybenefits. Analysts build their skills andremain professionally relevant, andyour company enhances its analyticalcapabilities. Further, our research showsthat developing analysts actually makesthem much more likely to stay withyour company, despite their increasedattractiveness to other employers.

Deploy your analytical talent

To meet analytical objectives, you needthe right people with the right skills inthe right places at the right times.When analytical skills are well matchedto the work, the business is more pro-ductive and analysts are more engaged.There are three important componentsto deploying analytical talent effectively.

Organizing analytical talentThe way your analysts are organizedaffects their ability to perform theirjobs, develop their skills and work withother parts of the business. It can alsoinfluence their level of engagement,their degree of investment in yourcompany’s success and their likelihoodof remaining loyal.

Most companies with a successfulanalytical approach have centralizedtheir top pros and semi-pros to somedegree. Procter & Gamble, for exam-ple, took analytical groups that hadbeen dispersed and combined them toform a new Global Analytics group aspart of an IT organization. Poolinganalytical talent—or at least networkingit more cohesively—can stretch thesescarce resources by making it easier tosupply skills, advice and solutions tocommon problems.



that they’re using. However, just as manyare the direct result of an analyst’sfailure to appreciate the business sideof the analyses; for example, usingthe right definition of “customer” intheir model.

Relationship and consulting skillsAnalytical people are often stereotypedas nerds who have no relationship skillsand probably don’t know—or care—much about the actual business. Notonly are such assertions unfair, theyare highly counterproductive. Forcompanies that really want to becomemore analytical, their pros and semi-pros simply must be able to talk inlayperson’s terms with clients, withcustomers and with executives acrossthe organization.

For this reason, consulting and rela-tionship skills are important. Analystsmust be able to conceive, specify,pilot and implement analytical appli-cations; they must know how to advise,negotiate and manage expectations.Analysts also need to be able tocommunicate the results of analyticalwork, either within the business toshare best practices and to emphasizethe value of analytical projects, oroutside the business, to shape workingrelationships with customers, suppliersand regulators.

Coaching and staff development skillsThe ability to nurture and developanalysts is vitally important. If youranalysts aren’t centralized, coaching canensure that best practices are sharedacross the company. Good coachingnot only builds all the essential skillsdescribed here, but it also helps peopleunderstand how data-driven insightscan create business value.

Business insightAnalytics doesn’t happen in a vacuum.In order for people to know whereand how best to apply their analyticalskills, they need to understand thestrategy for their business, functionand even department. They also needto understand the organization’s corecapabilities and how analytics cancreate value for the business.

One way to increase your analysts’business acumen is through develop-mental assignments. Be sure to exposeyour analysts to the various businessunits and functional areas acrossthe company so they learn about thecompany’s main business challengesand work processes. E&J Gallo Wineryrotates its analytical pros betweendifferent business units and functionaldepartments during 18- to 24-monthtours of duty. As they analyze grapesupply, develop new customer seg-mentation models or perform supplieranalyses, these quant jocks becomesavvy at identifying opportunities foranalytics to help the business.

Semi-pros (and even amateurs) needto continually hone their businessknowledge and skills to avoid thesometimes costly errors that are sur-prisingly common. (Experts estimatethat between 20 percent and 40 percentof all spreadsheets have errors—somecatastrophic.9 ) Many analytical errorsare the result of employees not under-standing the data or the technology



There’s no single organizational modelthat’s best for every company. Theright model will depend on the matu-rity of your company’s analytics func-tion and whether you need analyststo be closer to one another or closer tothe business. Nevertheless, our researchreveals that firms with a strong com-mitment to building an analyticalworkforce are best served by greatercentralization and coordinationof analysts.

If you’re new to analytics, a first stepmight be to create a community ofinterest to provide support and shareexamples of best practices. Onceyou’ve built a critical mass of expertsand demand for analytics grows, youcan look at other organizational mod-els, such as a centralized analyticsfunction or a center of excellence.(We discuss this topic in more detailin the next section, “How to OrganizeYour Analytical Talent.”)

Clarifying analysts’ career pathsIn most organizations, pros and semi-pros are measured and developed notaccording to their analytical skills butaccording to the standards of thebusiness units or functions in whichthey are housed. As a result, rewards,performance management, objectivesand role descriptions often vary wildlyfor analytical talent in the sameenterprise. Creating distinct roles canboost performance and engagementby ensuring that analysts have clearlydefined objectives, reward structuresand growth opportunities.

Creating unique career paths can alsoincrease your ability to attract andretain top analysts. Jim Heffernan, CFOof the Massachusetts General PhysiciansOrganization, says that it’s a big chal-lenge to avoid setting analysts up forfailure by promoting them accordingto the organization’s standard careermodels. Many of the most technicallyadept analysts don’t want to manageothers. To ensure that it can retainthese valuable analysts, the organiza-tion is creating a compensation andpromotion system that will allow greatanalysts to be individual contributorswithout direct reports and still be eli-gible for pay increases and promotions.

Staffing analytical talentMatching the right pros and semi-prosto the right project is critical. Be pre-pared to move people around in orderto help them focus on the company’sbiggest problems.

The prerequisite to appropriate rede-ployment: An enterprise-level perspec-tive on analytical roles and a detailedinventory of the organization’s analyt-ical skills. When Best Buy found itselfwith many open analytical positions,leaders evaluated the skills availableacross the business. A review of skillsand open positions found many ana-lysts who were either under- or over-qualified for their work, so the compa-ny redeployed people to new positions,putting hundreds of employees’ skillsto better use while greatly improvingemployee job satisfaction andengagement.

Managing analytical talentstrategically

Analytical talent is vital to the successof nearly every organization. Whetheryour company aspires to compete onanalytics or simply to become moreanalytical, you need to manage youranalytical talent as a strategic workforce.

This cannot be accomplished withpiecemeal, ad hoc approaches that touchonly on parts of the overall talentmanagement equation. It requires anapproach that looks across the entireenterprise. By building and aligningthe four key talent management capa-bilities, companies can maximize thestrategic impact of their analyticaltalent and continually expand theorganization’s collective analyticalcapabilities. Or, in simpler terms, theycan build their own talent-poweredanalytical organization.



How to Organize YourAnalytical Talent



To get the most value from your analytical talent whilealso keeping them engaged and on board, you have toorganize them with the right approach. We discovered thatcompanies typically employ one of five models to organizetheir analytical talent. Our research shows which models aremore effective both for the business and for the individuals.The key? Keep your analysts both “close to the business”and “close to each other.”

Figure 2: Options for organizing analytical talent

Centralized

Corporate Corporate Corporate Corporate Corporate

Center of excellence Consulting Functional Decentralized

Businessunit

Businessunit

Businessunit

Business unit Business unit

Analytics group

Analytics project

Function FunctionCOE Function Function Function

Solid line indicates a direct line of authority

Dotted line indicates a partial line of authority or funding

Large, multi-divisional corporations use one of five models to organize analytical talent.

Business leaders know they rely onanalytical talent to maintain theircompany’s competitive edge. But theyoften struggle with one crucial question:What’s the best way to organize ouranalysts? Executives have to decidewhether they should be centralized ordecentralized; “charged out” to the restof the business as consultants or madeavailable as a free resource; and whereand to whom they should report.

The stakes of using the right model arehigh. When companies fall short, theirbest analytical minds are relegated toconducting simple analyses or workingon low-value projects rather than build-ing robust models to solve the mostchallenging business problems. Worsestill, that type of work is a sure recipefor analyst disengagement and defection.

centralization and coordination ofanalysts. The right organization modelis also critical to engaging and retain-ing this ever more important segmentof the workforce.

To decide which model is right foryour situation, you need to start byasking questions about your level ofanalytical maturity and the speed andlevel of change your organization isprepared to undertake.

How companies organizeanalytical talent

Based on our research and work withanalytical organizations, we identifiedfive basic options for organizinganalysts in large, multi-divisionalcorporations. (See Figure 2.)

But companies seeking to maximizeanalytics’ impact on the business whilebuilding a highly engaged analyticalworkforce face a dilemma. They needto organize analysts in a way that hasthem working “close to the business”on the most important analyticalinitiatives while keeping them working“close to one another” to coordinatetheir efforts and for purposes of mutuallearning and support. Making bothhappen at once is the challenge.

Are companies up to the task? To findout, we interviewed dozens of executivesand surveyed more than 700 analysts.We discovered that companies use oneof five models to organize their analysts.While the right choice can vary, wefound that companies with a strongcommitment to building an analyticalworkforce are best served by greater



Centralized. Analysts reside in onecentral group where they serve avariety of functions and business unitsand work on diverse projects.

Besides providing analytical expertiseand support, this unit sets the analyti-cal direction of the entire organization.This structure makes it easier to deployanalysts to projects with strategicpriority. Candy giant Mars uses thisapproach; its Catalyst group has long-term funding and works on strategicinitiatives across the business. Onecaveat: This model can create distancebetween analysts and the business,particularly if all analysts are housedin one location.

Center of excellence. Analysts areallocated to units throughout theorganization, but their activities arecoordinated by a central entity.

The center serves to build a communityof analysts, primarily to share knowl-edge and best practices with oneanother. It can also double as a projectmanagement office, looking acrossanalytical initiatives and determiningproject priorities and staffing. CapitalOne and Wal-Mart both employversions of this model. At Capital One,Ph.D.-trained statisticians reside in acenter of excellence. At Wal-Mart,the Information Systems Division hasorganized an analytics center ofexcellence, even though other divi-sions employ most of the analysts.

Consulting. Analysts work togetherin a central group but act as internalconsultants who charge “clients”(business units) for their services.

The consolidation of analysts enablesenterprise-wide coordination ofanalytical activities. However, a lackof direction from senior leadership cancause problems because top analystscan end up working on small, peripheralproblems rather than key strategicchallenges. United Airlines, eBay, andtrucking firm Schneider National allemploy the consulting approach.

Functional. Analysts are located inthe functions where the most analyticalactivity takes place, such as marketingand the supply chain.

In some cases, all analysts are housedwithin the function with the strongestanalytical orientation, even thoughsome may do work for other parts ofthe business. Fidelity uses this approach:The great majority of analysts workin the Customer Knowledge group,which reports to marketing. Analystsare found in large numbers inmarketing at Carnival Cruise Lines andat GE Money (although they also workfor the risk management functionand other groups).

Decentralized. Analysts are scatteredacross the organization in differentfunctions and business units with littleto no coordination.

This model makes it difficult to setenterprise-wide analytical prioritiesand to develop and deploy staffeffectively. Companies with fewanalysts and little management supportare most likely to be decentralized.

Our survey data reflect the variety oforganizational models across industries.The decentralized model is mostprevalent (reported by 42 percentof analysts surveyed), indicating theimmaturity of most corporations’ ana-lytical capabilities today. Consumer-oriented industries that we’d expect tohave advanced analytical capabilities—including retail, consumer goods andservices, financial services and healthand life sciences—tend to use morecentralized models. Company size wasn’ta key factor in choice of model, exceptin the most extreme cases: Companieswith fewer than 1,000 employees aremore likely to organize their analystsinto one functional area, whereascompanies with more than 25,000employees are more likely to use acenter of excellence model.

Analysts who work together incentralized groups or are connectedthrough a center of excellence reportedthe highest levels of engagement andare the most likely to stay with theircompanies. In fact, analysts in central-ized groups were nearly twice as likelyto be highly engaged as analysts whowere decentralized. (See Figure 3.)Analysts in centers of excellencereported the strongest intentions tostay with their employer, whereasthose in the functional, consulting anddecentralized models reported signifi-cantly lower levels of engagement andweaker intentions to stay. (See Figure 4.)





Figure 5: Key features and priorities of the main approaches to organizing analytical talentCentralized units and centers of excellence outperform other organization models on coordinating analytics initiatives,sharing and developing analysts’ knowledge, and deploying analysts strategically.

Type FeaturesCoordinateenterprise-wideanalytics initiatives

Centralized • Organized in a single,centralized unit

• Sets analytical directionof entire organization

• Expedia• Mars “Catalyst”

Group

Center ofexcellence

• Analysts decentralized,residing in business unitsor functions

• All analysts members ofcorporate center of excellence

• Capital One• Wal-Mart

Consulting • All analysts are part of oneorganization

• Business units “hire” analysts

• eBay• Schneider National• United Airlines

Concentratedin onefunctional area

• Analysts organized in asingle business unit orfunctional area—usuallyfinance or marketing

• Fidelity• Carnival Cruise Lines

Decentralized • Analysts sit in businessunits or functions

• There is no corporateconsolidating structure

• Majority ofnon-analyticalcompanies

Deploy analystsstrategically

Buildrelationshipswith the business

Share bestpractice &training

Company examples

Priorities

Low High

Centralized

Center of excellence

Internal consultants

Concentrated in one functional area

Decentralized

33%

28%

41%

22%

24%

33%

30%

27%

27%

29%

Strong intentions to stay(top quartile)

Weak intentions to stay(bottom quartile)

0 5 10 15 20 25 30 35 40 45

Figure 4: Analyst intentions to stay in the five organization modelsAnalysts in centers of excellence are most likely to intend tostay with their company and least likely to intend to leave.

Centralized




Decentralized

35%

17%

29%

14%

23%

28%

24%

24%

18%

24%

Highly engaged (top quartile) Disengaged (bottom quartile)

0 5 10 15 20 25 30 35 40 45

Figure 3: Analyst engagement in the five organization modelsAnalysts in centralized units and centers of excellence are mostlikely to be highly engaged and least likely to be disengaged.


Centralized




Decentralized

39%

18%

29%

18%

26%

22%

25%

17%

22%

32%

High meaningful opportunities (top quartile) Low meaningful opportunities (bottom quartile)

0 5 10 15 20 25 30 35 40 45

Figure 6: Meaningful work and career opportunities in the five organization modelsAnalysts in centralized units are most likely to report that they have meaningfulwork and career opportunities. They use their highly specialized skills and gainvaluable experience while contributing to the company’s goals.

Three key factors influence the qualityof analysts’ work and career opportu-nities (and, in turn, drive engagementand retention): that their work isaligned with the organization’s strategyand goals and affects its success (sig-nificant work), that they understandthe business (business insight) andthat their skills and aspirations are agood match with the company cultureand goals (organizational fit).

Significant workWithout the chance to make a realimpact on the organization’s success,analysts won’t find enough meaningin their work, and so they will beless engaged and less likely to stay.Perhaps the biggest de-motivator foranalytical pros is spending too muchtime on simple analyses and reportgeneration instead of building and

refining analytical models. We knowof several organizations that have lostanalysts who felt they were treatedlargely as “spreadsheet developers.”It’s essential to give your best analystsopportunities to apply their expertiseto the company’s biggest problems.

One analyst, Sharon Frazee, gave usan example of how significant workwas connected to engagement. At onepoint in her career, she led a successfuland highly engaging initiative to stan-dardize information and to streamlineits distribution—changes that gave linemanagers more time to analyze potentialgrowth opportunities. Frazee, nowvice president of corporate healthcareanalysis and research at Walgreen’s,


If engaging analysts were the solegoal, our recommendation would besimple: Centralize them. But in practicewe’ve found that the ideal modeldepends on a company’s priorities, thematurity of its analytical capabilitiesand the need to balance analystsupply and demand. (See Figure 5.)When just starting to use analytics,most companies don’t need to usethem in every area of operations. Nordo they have enough analysts to justifycentralizing resources—making a func-tional model a natural fit. As demandgrows, companies tend to hire moreanalysts. Once an organization getsa critical mass of experts and demandreaches the point that allocation ofthese scarce resources becomes apriority, one of the more centralizedmodels is appropriate.

How organization affectsanalysts

Meaningful work and career opportu-nities are critical for engaging andretaining all types of employees, andanalysts are no exception. How com-panies organize analytical talentaffects whether they have access tothe most meaningful opportunities—work that allows them to use theirhighly specialized skills, gain valuableexperience and contribute to theorganization’s goals. Analysts in cen-tralized units and centers of excellenceare most engaged and most likelyto stay because they enjoy the mostmeaningful career opportunities.(See Figure 6.)

36%

18%

28%

20%

23%

27%

30%

20%

22%

24%

High business insight (top quartile) Low business insight (bottom quartile)

0 5 10 15 20 25 30 35 40 45

Centralized




Decentralized

Figure 7: Significance of the work in the five organization modelsAnalysts in centralized units and centers of excellence were most likely to say thatthey do important work that has a significant impact on their company’s success.

Centralized




Decentralized

40%

26%

38%

19%

32%

34%

28%

30%

32%

27%

0 5 10 15 20 25 30 35 40 45

High significance (top quartile) Low significance (bottom quartile)

Figure 8: Analysts’ business insight in the five organization modelsAnalysts in centralized units reported the greatest understanding of theircompany’s strategy and capabilities, lines of business and competitive environment.Because of their close ties to the business, analysts in functional groups alsopossess considerable business insight.

described her passion to contribute:“I want to make a difference. Being ableto do the kind of informatics work thatactually gets applied, and to seethings changed because of it, isimportant to me.”

How analysts are organized greatlyaffects their opportunities to makemeaningful contributions. Companiescreate centralized units and centersof excellence to ensure that analystsare working on the models and appli-cations that matter most to the business.Indeed, we found that analysts incenters of excellence and centralizedunits were most likely to report thatthe work they do is important. (SeeFigure 7.) Analysts who are organizedas internal consultants, housed infunctional units or dispersed throughoutthe business were more likely to bestuck working on peripheral projects.

Business insightOrganizations don’t just want “quantgeeks.” They need highly quantitativepeople who understand the businessand can develop strong relationshipswith business leaders. How analystsare organized can have a major impacton this dimension.

According to our survey, analysts incentralized units, functional groupsand centers of excellence had moreinsight into their company’s businessthan analysts in internal consultinggroups or analysts who were decen-tralized. (See Figure 8.) These more





centralized groups offer analysts anenterprise perspective: One in threeanalysts in centralized units under-stands how his or her group con-tributes to the organization’s success,whereas just one in five analysts inthe other models has this degree ofinsight. (See Figure 9.)

GE is famous for its ability to buildbusiness acumen far and wide through-out the organization. It’s no differentwith the company’s analytical talent.For example, its financial services unit,GE Money, operates analytical centersin Shanghai and Bangalore. Staff mem-bers from the centers routinely rotatethrough other parts of the business. Theseassignments help analysts learn aboutthe local operation and better preparethem to meet its needs through thecenter in the future. Equally important,the assignments are a valuable reten-tion tool because they offer analysts asense that they are making meaningfulcontributions to the business.

Organizational fitIn many companies, analysts are vieweddifferently from the rest of the work-force; one executive we interviewedcalled them a “weird species.” Analyststhemselves take pride in their uniqueness.Still, they want to work for companiesthat value analytics and with colleaguesthat appreciate and respect theirtalents. They are most engaged whenthey believe their rare and valuable skillsare a good match with the company’sculture and goals.

Figure 9: Line of sight to the business in the five organization modelsAnalysts in centralized units were most likely to understand how their groupcontributes to their company’s success.

Centralized



Decentralized


33%

17%

21%

28%

20%

36%

19%

33%

19%

26%

High line of sight (top quartile) Low line of sight (bottom quartile)

0 5 10 15 20 25 30 35 40 45

Centralized



Decentralized


47%

11%

30%

17%

26%

19%

43%

14%

27%

19%

High organizational fit (top third) Low organizational fit (bottom third)

0 5 10 15 20 25 30 35 40 45 50

Figure 10: Organizational fit in the five organization modelsAnalysts in centralized units were most likely to report a high degree of fit withtheir company.

We found that organizations witha centralized analytics function arebetter at communicating the valuethey place on analytics and at makinganalysts feel at home. Nearly half ofall analysts working in centralizedgroups report a high degree of fit withtheir organization; only one in 10 feelstheir company is not a great fit forthem. (See Figure 10.) The fact thatthey’ve chosen to bring together thecompany’s best analytical minds islikely a testament to the importancethey place on analytics. By contrast,analytical talent deployed as consul-tants or concentrated in one functionalarea often feel like misfits, isolatedfrom the business as well as fromother analysts.

Grouping analysts together, whetherin centralized units or centers ofexcellence, helps to maximize the fitbetween analysts and the organizationand thus keep analysts engaged. Forexample, Dr. Steven Udvarhelyi, seniorvice president and chief medical officerat Independence Blue Cross, reportsthat the company’s internal “informat-ics organization” is critical for devel-oping analytical talent: “It’s a definedcenter of excellence. It creates a fer-tile ground for people to work withother people. It creates a critical masswhere you get career opportunities,growth opportunities and goodprofessional interaction.”

Getting organized

How do you determine which model isbest for you? That answer dependsupon the maturity of your company’sanalytical capabilities as well as thelevel of demand for analytical skills.If your organization has little demandfor analytics, has only a handful ofanalysts and no executive sponsorship,then it’s best to house analysts whereverthey have a willing sponsor. In moreanalytically sophisticated companies,demand for analytics quickly outstripsthe supply. In this case a hybridapproach, with a centralized team ofthe most advanced analysts comple-mented by a center of excellence foranalysts deployed to functions or busi-ness units, may be the best solution.An enterprise-level organizational modelmakes it much easier to deploy analystson strategic business priorities.

The next challenges: How fast canyou get there, and what’s the idealpace of change? Organizational changeshould be segmented into stages andspread over multiple time horizons—quick wins versus longer-term goals.For example, linking decentralizedanalysts through a community ofinterest is relatively easy to do. Settingup a formal center of excellence, withspecialized training and developmentfor its members, takes longer. Thepotential for change will be determinedby the maturity of your organization’sanalytics function as well as the strengthof leadership and the level of seniorexecutive commitment to analytics.

What if you can’t easily change yourorganization’s model? Leave analystswherever they have enough sponsorshipto protect them and advocate foradopting a more enterprise-wideperspective (which would require,for example, consistency in skills, jobdescriptions, training, salary andcareer development opportunities).If you use the consulting model, findways to demonstrate to analysts howtheir work contributes to organizationalgoals. It’s also important to build asense of community in order to sharebest practices and improve analysts’sense of fit with the organization. Ifyour analysts are concentrated in onefunctional area or are decentralized,you’ll need to find ways to providethem with an enterprise perspectiveand to build links between them andwith other parts of the business. Buildingan informal community of interestmay be the best place to start.

And if your analysts are alreadycentrally organized or in a center ofexcellence, that doesn’t mean all yourproblems have been solved. Whilecentralized organizational structuresare superior, they’re not perfect. Exploreways to link centralized analysts moreclosely to people in the business sothat they better understand their needsand can more effectively translate thefindings of their analyses. If you havea center of excellence, it’s importantto reinforce the links between analystsand to make sure that senior managementfully supports the analytics function.



Beyond organization

Companies and managers need torecognize analytical talent as a specialsegment of high-value employees,whose preferences and motivationsvary significantly from those of otheremployees. Organizing analytical talentin a way that not only addresses thestrategic and operational needs of thebusiness but also provides analystswith meaningful work and careeropportunities is essential. (See Figure 11.)Regardless of which model a companyadopts, it should develop a communityof analysts and take an enterpriseapproach to organizing them. That’s thebest way to ensure analysts feel thatthey fit with the organization, under-stand the business and have opportuni-ties to make meaningful contributionsto its success.

The significance of analysts’ work, theirinsight into the business and their fitwith the organization will all be higherin “analytics friendly” organizations.Managers that value fact-based analysiscreate precisely the kind of challeng-ing and important analytical assignmentsthat make analysts thrive. When man-agers don’t value analysts’ work, thenno matter how they are organized, theywill be underused and unappreciated—two guaranteed ways to sap themotivation of the best analysts andto send them to your competitors.



Figure 11: Key performance indicators for the five organization modelsThe centralized and center of excellence models outperformed the others across a range of dimensions.

Significant work

Centralized

Center ofexcellence

Internalconsultants

Decentralized

Business insight Organizational fit Meaningful work &career opportunities

ConditionsInfluences Outcomes

Engagement Intentions to stay

Concentratedin onefunctional area

Low High



How to Engage andRetain Your Analytical Talent



What makes analysts tick? Knowing the answer iscritical if you want your analytical talent to fully investthemselves in their work and the company’s success.Our research reveals four things companies mustdo well in order to engage and retain this scarce andvaluable breed.

What Is Engagement?

Engaged employees invest their physical, mental and emotional energies into their organization and its success. We gaugedrespondents’ engagement levels by measuring the reported frequency with which they displayed behavioral, intellectual oremotional investment in their work. Possible responses ranged from “Never” (1) to “Always, every day” (7). The data revealthat, on average, analytical talent is more likely to be engaged than other employees.

Analytical Non-analyticalEngagement survey item talent talent

I am enthusiastic about providing a high-quality product or service. 52% 39%

I am determined to be complete and thorough in all my job duties. 50% 41%

I am always willing to "go the extra mile" in order to do my job well. 43% 33%

I am prepared to fully devote myself to performing my job duties. 35% 27%

My job is a source of personal pride. 33% 24%

I am willing to really push myself to reach challenging work goals. 32% 26%

Trying to constantly improve my job performance is very important to me. 29% 23%

I am ready to put my heart and soul into my work. 28% 24%

I get excited thinking about new ways to do my job more effectively. 20% 13%

Percentage responding “Always, every day”

Although the business literature isrife with studies of how to engageand retain employees, analyticaltalent is unique. The analysts, “mathbrainiacs” and “Excel ninjas” havedistinct backgrounds, skills, attitudesand motivations. To complicate mattersfurther, they themselves are a diversebunch–from the executive championswho lead major analytical initiatives tothe professionals who build and applystatistical models and algorithms tothe many employees who regularlyuse data and analytics in their work.Managers must understand whatmotivates these workers in order tosuccessfully engage and retain them.

To discover the most important factorsfor engaging and retaining analysts,we interviewed dozens of executivesand surveyed 1,367 U.S.-basedemployees, including 799 analysts.(For additional details, see the appendix,“About the Research.”) We examinedmore than 30 factors known to predictemployees’ levels of engagement andintentions to stay with their employer,such as company culture, organiza-tional systems, management practices,job and career opportunities, leader-ship and management and co-workerrelationships.10

We determined that the essentialsof engaging and retaining employeeshold true: Companies need toprovide meaningful work and careeropportunities, support people’s effortsto engage as well as their need for

recovery and renewal and cultivate aculture of trust and respect.11 But ourresearch revealed something important:Analysts have a different view onwhat makes opportunities meaningful,what kinds of support are essentialand what cultural factors matter most.



Champion Professional Semi-professional Amateur Non-analytical talent

Disengaged Somewhat engaged Moderately engaged Highly engaged

8.2%10.5

21.422.7

32.3

12.3

20.9

25 24.4 23

45.2

33.7

20.9

26.3

21.4

34.234.932.7

26.6

23.3

Figure 12: Level of engagement by type of employee

On average, analysts are more engaged than non-analysts. Analytical champions and pros are much more likely to be highlyengaged than other employees. Among analysts, semi-pros and amateurs are most likely to be disengaged.



The importance of engagingand retaining analytical talent

When employees are engaged, theyinvest their full physical, mental andemotional energies in their work.Engaged employees go beyond whatis expected or required—they showinitiative, innovate and continuallyfind new ways to contribute to theirorganization’s success. Employeeengagement is one of the keys to highperformance. Various studies haveshown that highly engaged workforcesproduce better business results, asmeasured in terms such as higherprofitability, productivity and customersatisfaction.12

The good news is that, according toour research, analysts as a group weresignificantly more engaged at workthan other types of employees. (Seethe sidebar, “What Is Engagement?”)Overall, 57 percent of analysts reportedbeing moderately or highly engaged,compared with 45 percent of otheremployees. (See Figure 12.) Analyticalwork may seem deadly dull to thenon-quants among us, but it possessesmany of the characteristics of moti-vating and engaging work.13 Whenanalytical work is well-organized andaligned with business needs, analystshave the opportunity to use a diversemix of skills (quantitative, technicaland interpersonal), see a project throughfrom start to finish and make a signif-icant impact on business outcomes.They also enjoy a degree of autonomyin performing the work, and they

receive instant feedback on theirperformance (either a model works orit doesn’t). Indeed, the analysts wesurveyed registered significantly higherlevels on these factors than otheremployees did.

The bad news, however, is that asizeable minority of analysts is muchless engaged. One in four was simplygoing through the motions—thesepeople show up for work each day, butthey don’t give their all. And a full 20percent of analysts were completelydisengaged—rarely, if ever, investingfocus or passion in their work or intheir company’s success.

A high attrition rate among analystsis a major problem that companiescan’t afford to ignore. When analystsjump ship, they take vast amounts ofknowledge and expertise with them.If highly specialized analytical profes-sionals leave, there is often no onewith the same depth of technical skillto replace them. What’s more, findinganalytical skills in the market is becom-ing more challenging as the supplyof new graduates shrinks and demandfor quantitative skills increases.16

Therefore, companies need a clearunderstanding of what engages ana-lysts and what motivates them to stayor leave.

In our research, we compared thefactors that influence engagement andintentions to stay among analysts withthose that matter to non-analyticaltalent. We discovered that several

search for a new job in the next yearis unsettling enough. But another 19percent said that they are “unsure”if they will stay with their employerin the coming year. That means halfof the analysts we surveyed are eitherunlikely to stay or ambivalent aboutstaying with their companies very long.

The data for analytical professionalsis most troubling. (See Figure 13.) Acompany’s Ph.D. statisticians, mathe-maticians and other quantitativespecialists may appear to be immersedin their statistical models, but manyare also updating their resumes andtaking calls from headhunters. Nearlyhalf (47 percent) of the analyticalprofessionals we surveyed said theyare likely to actively look for a new jobin the next year. Many semi-profes-sionals don’t plan on sticking around,either: 33 percent expect to start ajob search.15

More troubling are our findings onretention. Even highly engaged analystswon’t necessarily stay at their currentorganizations. Too many executivesmistakenly equate employee engage-ment with retention, but employers’best efforts to engage analysts maynot keep them from walking out thedoor.14 Our research reveals thatengagement level accounts for lessthan 10 percent of a person’s inten-tion to stay. Therefore, as well asengaging analytical talent, managersneed to figure out how to retain them.

The best analytical talent has noshortage of opportunities: Companiesfighting to retain analysts face anuphill battle. The fact that 31 percentof analysts responded that it is “likely”or “very likely” that they will actively



Figure 13: Likelihood that respondents will actively look for a new job in the next year

On average, analysts were less likely to intend to stay with their current employer than non-analysts. The more specializedand valuable their skills, the more likely analysts are to consider other options—champions, pros and semi-pros were mostlikely to say they will soon look for a new job.

Champion Professional Semi-professional Amateur Non-analytical talent

Not at all likely Unlikely Unsure Likely Very likely

12.3%11.6

20.9

35.2

28.126

18.620.5

24.922.5

19.2

23.325.5

15.1

20.9 20.521.9 20.9

16.8

12.213

25.6

16.4

12.7

15.6

Figure 14: What matters most for engaging and retaining analytical talent

Business insight

Influences Conditions Outcomes

Meaningful work &career opportunities

Resources to supportengagement & renewal

Role clarity

Supportivemanagement

Culture of trust& respect

Engagement

Intentions to stay

Technical skills

There are three essential conditionsfor engaging and retaining analystsand non-analysts alike: Companiesneed to provide meaningful work andcareer opportunities, support people’sefforts to engage as well as theirneed for recovery and renewal andcultivate a culture of trust andrespect. But some things are uniquelyimportant to analysts. Companiesmust do four things well to engageand retain analytical talent: developtheir business insight, providerole clarity, help them expand theirtechnical skills, and ensure manage-ment support.



factors that have very little effecton other employees’ engagementor intentions to stay are crucial forengaging and retaining analyticaltalent. Companies must do fourthings well to engage and retainanalytical talent:

1. Arm your analysts with criticalinformation about the business.

2. Set roles and expectations:for analysts, clarity is critical.

3. Feed your analysts’ love of newtechniques, tools and technologies.

4. Give analysts the managementsupport they need.

These are especially crucial for engag-ing and retaining your most quantita-tively savvy analytical talent—theprofessionals who create and use thecomplex models and algorithms thatcompanies rely on to inform businessdecisions.

How to engage analytical talent

Meaningful work and career opportu-nities are critical for engaging alltypes of employees, and analystsare no exception. Analysts are mostengaged by work that allows them touse their highly specialized skills, gainvaluable experience and contributeto the organization’s goals. Two keyfactors influence the quality of analysts’work and career opportunities (and,in turn, drive engagement): Businessinsight, or how well they understandthe business, and role clarity, or howfully they understand their role.

In addition, employees are most engagedwhen they have the resources they needto be effective—the skills, information,people, technology and tools thatmake it possible to do their jobs well.The opportunity to continually updatetechnical skills is uniquely importantto analytical talent. (See Figure 14.)

Arm your analysts with criticalinformation about the businessAs analytics becomes more integralto a company’s strategy, analysts needto develop stronger business insight.They can’t spend their days with theirheads buried in models and spread-sheets. They need the businessknowledge and skills to allow them tounderstand the strategic issues facingthe company and how analytics canbe used to drive business value.Insight into the business not onlymakes analysts more effective—it alsoboosts their engagement.

Our research revealed that businessinsight is one of the strongest predictorsof analyst engagement. More thanhalf of the analysts who understandtheir company’s strategy, goals, capa-bilities and operations were highlyengaged, compared with just 18 percentwho don’t have a firm grasp of thebusiness.17 (See Figure 15.) Moreover,analysts who understand how theirwork relates to their organization'sgoals and contributes to its successwere nearly six times more likely to behighly engaged than those who don’t.

Business insight is particularly impor-tant when it comes to engaginganalytical professionals. Despite theirtechnical background, these pros aremost engaged when they thoroughlyunderstand how their models and

analyses can be used to good effect.In our survey, analytical professionalswith considerable appreciation for thebusiness were twice as likely to behighly engaged.

To develop analysts’ business insight,companies should expose them to arange of business units and functionsso they learn about the company’smain business challenges and workprocesses. Leaders at a global financialservices company stress the need foranalysts to understand the businessside of things and to engage withexecutives on their terms: “Let thembe executives. Discuss business issuesand the potential for analytics tohave an impact on the organization’sresults,” says the head of one of thecompany’s analytics groups. The firmgives analysts the tools and templatesthey need to capture business strategy,



delineate problems and devise solu-tions. Analysts there can effectivelycommunicate and negotiate withbusiness leaders because they canexplain how their work creates valuefor the firm.

Opportunities to interact with keystakeholders within and outside thecompany help analysts learn how thebusiness uses analytics and how tocommunicate the results of theiranalyses in non-technical language.Analysts who have the knowledgeand skills necessary to communicateeffectively with suppliers, customersand other stakeholders are six timesmore likely to be highly engaged,we found.

Colin Sheppard, Virgin Media’sDirector of Knowledge and Insight,says Virgin trains its analysts to thinklike clients. It encourages them tofocus on the business’s most pressingproblems, not simply to have fun withthe data. He finds that’s what makesthe difference between good andgreat analysts—the best are not onlytechnically outstanding, they alsounderstand the key motivations of theconsumer and are focused on com-mercial objectives. Analysts who canconfidently communicate their findingsin terms that are important to seniorexecutives are not only more engagedbut also more likely to convincemanagement to act upon their recom-mendations.

Figure 15: Percentage of respondents that are highly engaged, given highand low levels of business insight

0 20 40 60 80 100

Analytical talent

Non-analytical talent

51%

18%

35%

18%

High business insight Low business insight

Insight into the business matters more for engaging analysts than non-analysts.Analysts who possess business insight are nearly three times more likely to behighly engaged than those who don’t.

0 10 20 30 40 50 60 70 80

High role clarity (top quartile) Low role clarity (bottom quartile)

Amateurs

Semi-professionals

Professionals

Champions

Non-analytical talent

59%

9%

64%

8%

75%

9%

68%

19%

57%

8%

Figure 16: Percentage of respondents that are highly engaged, given highand low levels of role clarity

Role clarity is essential for engaging analytical talent—especially analytical profes-sionals. Pros who have clear goals, objectives and expectations for their jobs aremore than eight times more likely to be highly engaged than those who don’t.

At Duke Energy, analysts must beprepared to represent the business andexplain its analytics to a wide range ofstakeholders—from internal customersto regulators and the public. A proposalto build a new power plant, for example,entails extensive analysis of an enor-mous amount of data. Analysts arecalled upon to publicly explain anddefend the company’s models. Accordingto Dick Stevie, Managing Director ofCustomer Marketing Analytics, thisrequires “precision, completeness andclarity—no jargon and no metaphor.You’ve got to convince a panel thatwhat you’ve done is logical and rea-sonable.” To prepare for these situa-tions, Stevie routinely puts analysts

in front of top management andgroup leaders for mock trials, wherethe analysts get practice testifyingand being cross-examined. Analystswho understand the external businesslandscape and can speak the languageof business are both effective andengaged.

Set roles and expectations: foranalysts, clarity is criticalAnalysts like structure. As a group,people with a strong quantitativeorientation tend to be less tolerantof uncertainty and think in a linearfashion.18 That’s why they are so goodat what they do—they turn raw datainto clear insights by creating modelsand applications that make sense of it.That tendency toward order leads ana-lysts to prefer structured and pre-dictable work environments.



It’s awfully frustrating when you don’tknow what you’re supposed to do.Engagement suffers in the absenceof clear goals and expectations—andthis is especially true for analyticaltalent. In fact, analysts who said theyunderstand their roles were six timesmore likely to be highly engaged. Theflip side? Analysts with ambiguousroles were nine times more likely tobe disengaged. A compelling argumentfor attention to this point if everthere was one.

At Google, employees know preciselywhat is expected of them. Roles arehighly structured according to a70/20/10 model in which employeesspend 70 percent of their time fulfill-ing basic job requirements, 20 percenton projects that help them developtechnical skills and benefit the company,and 10 percent on product and businessinnovations. Although aspects of therole are open-ended, overall expecta-tions and job requirements are clearlydefined. The company makes sure jobdescriptions are clear because, accord-ing to Liane Hornsey, an HR directorthere: “Good people only fail if theydo not know their role.”19 In addition,a set of 25 performance metricskeep “Googlers” on track to achievetheir goals.20

We found that role clarity is particu-larly important for engaging themost quantitative-minded analysts.Analytical professionals are muchmore likely to be engaged whenthey have a clear understanding oftheir responsibilities, objectives andauthority: Three out of four analyticalprofessionals who know what is expectedof them were highly engaged, whilejust one in ten pros who lack suchclarity was. (See Figure 16.)

Clear does not mean rote, however.Companies must be mindful thatanalysts place a premium on interest-ing and challenging work. In particular,pros and semi-pros want to work witha variety of datasets and types ofanalyses. One grocery retailer discov-ered this the hard way. The companycould attract highly skilled MBAsto a job that entailed an essential butrepetitive analytical task, but it couldnot keep them for long. The analystsquickly became restless and soughtnew challenges. Variety in their workand a sense of personal progress keepanalysts challenged and engaged.



In addition, the chance to work withthe latest analytic models, tools andtechnologies is crucial.

Feed your analysts’ love of newtechniques, tools and technologiesAnalytical work requires specializedskills, and the skill requirementschange rapidly as new analytical toolsand techniques emerge. To sustainand expand your organization’s ana-lytical capability, you must keep youranalysts’ technical skills up to date.Investing time and resources in main-taining your analysts’ technical skillswill also yield engagement benefits.In our research, we found that analystswho have opportunities to keep upwith advances and developmentsin their field were three times morelikely to be highly engaged than thosewho don’t.

Professionals, in particular, are muchmore likely to be engaged when theycan maintain their technical skills:Those who said they were able tokeep up with the latest tools andtechnologies in their field were fourtimes more likely to be engaged and26 times more likely to be highlyengaged. (See Figure 17.)

Consider the statisticians at AT&T Labs.The mandate of this analytical talent“is to develop new methodologies todeal with large-scale data problems—the type of problems generated by themassive stores of data AT&T collects torun its business,” says Chris Volinsky,director of the Statistics ResearchDepartment. To do this, it’s essentialthat they keep up with latest advancesin statistical theory and methodology.

Figure 17: Engagement levels of analytical professionals, given time andresources to keep up with field

Highly engaged

Moderately engaged

Somewhat engaged

Disengaged

0 20 40 60 80 100

Agree Neither agree nor disagree Disagree

79%

17%

3%

67%

17%

17%

50%

28%

22%

33%

44%

22%

Nearly four out of five highly engaged pros report that they are able to stayabreast of evolving tools and technologies.

making meaningful contributionsto their organization’s success, theyare likely to disengage.

Many semi-pros struggle to dividetheir time or to understand what theirjob responsibilities include and what isexpected of them. Clear requirementsand objectives sometimes don’t existfor their jobs. Semi-pros had signifi-cantly less clarity about their roles thanchampions, professionals or amateursdid. They work at the intersection ofanalytics and the business, often notknowing which way to turn first. Nowonder they are so stressed out.

To make matters worse, compared withthe other types of analytical talent,semi-pros receive the least amount ofsupport from their managers: Semi-pros need their managers to under-stand their problems and help themfind solutions. They also want theirsupervisors to recognize their potentialand encourage them to advance theircareers. But too many semi-pros toldus they don’t receive the support fromtheir managers that they crave. This isa surefire way to lose them.

When you walk the floors of yourorganization, give your semi-pros someextra attention. Their roles are fuzzy;they often lack adequate insight intothe business; and they don’t getenough support from their supervisors.But these beleaguered analysts areessential to your company’s success.

The Beleaguered Semi-Pro

All types of analytical talent demandspecial efforts to engage and retain,but companies should pay particularattention to the analytical semi-profes-sionals that apply analytics to businessproblems. Semi-pros lead difficult lives.They provide the interface betweenanalytics and the rest of the business.They sift through piles of data andprovide business solutions to demandingexecutives. They are central to theexecution of your company’s analyticsstrategy. Because they have a foot inboth worlds, their needs sometimes slipbetween the cracks. Out of all the typesof analytical talent, semi-pros are theleast engaged.

Even though their main job is to applyanalytics to business problems, semi-prosoften lack adequate insight into thebusiness, according to our research.In order to apply analytics effectively,semi-pros must understand their orga-nization’s goals, objectives and corecapabilities. They need to be aware ofthe external business landscape. Theymust recognize the key concerns ofvarious business departments. Despiteworking more closely to the business,many semi-pros reported that they don’thave a deep enough understanding ofit. Without that insight, and thereforewithout the sense that they are





The company expects them to be activein their disciplinary fields by publishingpapers, going to conferences, organiz-ing workshops and collaborating withcolleagues in other organizations.

These analytical professionals alsoexpand their technical skills by pursu-ing problems across the business andbeyond: “We have a lot of freedomto work on things that we find areinteresting to us,” says Volinsky. Hisgroup has done work in the areas oftargeted marketing, fraud detectionand manufacturing, for example.

A few years ago, the group took on achallenge posed by Netflix, the onlineDVD-rental company—and won. In2006, Netflix launched a competitionthat offered a top prize of US$1 millionto anyone that could improve—by atleast 10 percent—the accuracy ofCinematch, its movie recommendationalgorithm. “When they announced thecompetition, I thought that it was agreat match for the research that wedo at AT&T,” says Volinsky. The startingpoint was the enormous Netflix datasetof real customers’ movie ratings.

Volinsky and an AT&T Labs colleagueeventually teamed up with five statis-ticians, machine-learning experts andcomputer engineers from outside AT&Tto win the competition—three yearsafter it began. “When we startedworking on it, it wasn’t obvious whatthe tie-in was to AT&T. Once we were

spending serious time on it, it becamemore imperative to justify why we wereworking on this to our management,make a case for it and clear the pathto continue working on it. I was ableto justify working on it.”

“The algorithms that we developed forthe Netflix prize have benefited ourresearch here, and there’s definitelybeen a lot of interest in applying thetechnology internally to many differentprojects,” says Volinsky. “But thatfreedom to start working on it in thefirst place was a function of the cul-ture that we have here.” That cultureallows AT&T to make sure its topquant talent is constantly expandingtheir technical skills—and to engageand retain world-class analytical talent.

Google also builds developmentopportunities right into the work role.Employees spend one day a week onprojects that will develop their technicalskills and benefit the company. Thisarrangement is very important forengaging and retaining employees.Says HR Director Stacy Sullivan, "Itmakes people feel the company valuesthe employees." The numbers bear thisout. Fortune magazine named Googlethe number one place to work in theUnited States in 2007 and 2008,21

and turnover at the company is below3 percent in an industry that averagesdouble digit turnover.22

When companies build their analysts’skills, knowledge and competencies inways that also expand the organization’scollective capabilities, everyone wins.Analysts remain professionally relevantand marketable, and the companybuilds its analytical capabilities andbolsters long-term competitiveness.

How to retain analytical talent

To keep hold of your best analysts,you need to do more than make surethey are engaged. Although it’s certainlyeasier to retain engaged employees(highly engaged analysts are twiceas likely to intend to stay), as we’vestated, engagement itself is not enough.Fully 45 percent of the highly engagedprofessionals we surveyed are at riskof leaving within a year. The statisticsfor semi-professionals aren’t muchbetter. Four out of 10 highly engagedsemi-professionals are thinking aboutjumping ship.

To counter this threat, companiesneed to create a culture of trust andrespect—one in which people aretrustworthy, behave predictably, andsupport one another. In our survey,employees of every stripe were fourtimes more likely to intend to staywhen they reported high levels of trustand respect at their company. Thenumbers are even higher for analyticaltalent. Analysts who reported workingin a strong culture of trust and respectwere seven times more likely tointend to stay.

Several things influence employees’perceptions of a trusting and respectfulworkplace culture (and, in turn, driveretention): a sense that the organiza-tion values their contribution and caresabout their well being; the presenceof formal and informal practicesand procedures that support openinteractions; and the belief that theycan rely on their co-workers. In short,people expect their employers to



Figure 18: Analysts’ intentions to stay relative to their perceptions of a cultureof trust and respect.

0 20 40 60 80 100

The people in my company value others’ unique skills and talents

Employees in my company are able to bring up problems and tough issues

It is not difficult to ask others for help in my company

In my company, employees are not rejected for being different

When someone in my company makes a mistake, it is not usually held against them

In my company employees are free to take risks

No one in my company would deliberately act in a way that undermines others’ efforts

79%

54%

78%

56%

78%

51%

78%

50%

61%

29%

53%

36%

50%

32%

High intentions to stay Low intentions to stay

Analysts who say that their employer creates a trusting, respectful and supportiveworkplace culture are more likely to stay.

respect them, to create safe andsupportive work environments and toensure that employees act ‘withintegrity, professionalism, clear motivesand fairness. (See Figure 18.)

One additional factor is uniquelyimportant for retaining analyticaltalent—a really good manager. It’scommon knowledge that supervisorsplay an essential role in engaging andretaining employees. We found that’sespecially true for analytical talent.

Analysts are people too: give themthe management support they needAnalysts have a tough lot. Viewed astechnical specialists, they are oftenisolated from their business colleagues(or worse, from other analysts), andthey can feel that their contributionsare overlooked or misunderstood.But top analysts are rare talent, nothuman calculators. They need to feelvalued and supported by their employer,and they are likely to depart for greenerpastures if they don’t.

Their immediate supervisor is themost important factor in an analyst’sdecision to stay or to go. In our study,analysts were three times more likelyto stay when they believe their super-visor acts with integrity, treats peoplefairly and helps employees to succeed.Trust in one’s supervisor was especiallyimportant to analytical professionals:Pros with high trust in their supervisorwere eight times more likely to intendto stay than pros with low trust.



An effective working relationshipwith a supervisor is also essential.Analysts whose managers help themsolve problems, offer support whenthey struggle and recognize theirpotential are far more likely to stay.(See Figure 19.) For instance, theanalysts we surveyed were eighttimes more likely to intend to staywhen their supervisors take the timeto understand their problems andneeds at work. In addition, analystswho believe their supervisor is com-mitted to their professional growthand development were seven timesmore likely to intend to stay.

A senior executive with a globalfinancial services company told usthat finding and developing managerswho can do those things well is hisbiggest challenge in his 650-personanalytics group. The managers notonly need to understand analyticaltechniques and be skilled at buildinggood relationships with their businesscustomers, but they must also beadept at coaching and developingstaff. He actively nurtures those man-agers—keeping up with them regularlyby phone and meeting with thempersonally four to five days a year.

In order to sustain a highly engagedanalytical workforce, you need to makesure the people managing analystsunderstand their role in retention.

Moreover, you need to invest indeveloping your supervisors’ manage-ment skills. Supervisors with strongcoaching skills and an ability to groomanalysts for their next career step areparticularly important if your companyhas a large or fast-growing pool ofanalytical talent.

Figure 19: Analysts’ intentions to stay relative to the quality of their relationshipswith their supervisor.

0 20 40 60 80 100

High quality

Low quality

57%

44%

19%

82%

High intentions to stay Low intentions to stay

Percentage of respondents

Analysts who have a high quality relationship with their supervisor are three timesmore likely to intend to stay with their company.



Counting on analytical talent

Analytical talent is vital to everyorganization that relies on data-driveninsights, and your ability to engageand retain analysts is essential toyour company’s success. If you arecommitted to engaging and retainingyour analysts, recognize them as aspecial segment of high-value employ-ees, whose needs vary significantlyfrom those of the average employee.While analysts are motivated by manyof the same things that drive othertypes of talent, they also have distinctpreferences and aspirations. Indeed,our research revealed that severalfactors that have a relatively small

effect on other employees’ engagementor intentions to stay are crucial forengaging and retaining analytical talent.

Analysts are most engaged when theyunderstand the business side of thingsas well as the analytics, when theyknow what is expected of them, andwhen they can keep their technicalskills and expertise current. They aremost likely to stay when they have ahigh degree of management support.But when we examine how well com-panies are, in fact, arming analystswith critical information about thebusiness, setting clear roles andexpectations, feeding analysts’ desireto keep up with the latest tools andtechniques, and giving analysts themanagement support they need, wefind significant room for improvementacross the board. (See Figure 20.)

A few leading companies have alreadyput these insights into action and aresuccessfully engaging and retainingtheir valuable analytical talent. Butmost companies have a lot of work todo. By focusing on these four factors,you can improve analyst engagementand retention, enabling you to maximizethe strategic impact of your analyticaltalent and expand your organization’sanalytical capabilities.

Figure 20: How are companies doing? Key performance indicators by type of analyst

Business insight Role clarity Technical skillsManagement

supportEngagement Intentions to stay

Very poor ExcellentFairPoor

Semi-professionals

Amateurs

Professionals

Champions

Good

Companies must improve their ability to engage and retain all types of analytical talent—especially the analytical professionalsand semi-professionals who create and use complex models and algorithms and the amateurs who rely on the output ofanalytical models to do their jobs—by giving them more insight into the business, clearly defining their roles, developingtheir technical skills and providing the management support they need.

This report is based on three majorstudies carried out by researchers atthe Accenture Institute for HighPerformance. Jeanne G. Harris led agroundbreaking study of how organi-zations use analytics and publishedthe results in a best-selling book(co-authored with Thomas H. Davenport)entitled Competing on Analytics: TheNew Science of Winning (HarvardBusiness School Press, 2007). The bookexplains how high-performance busi-nesses are building competitive strate-gies around data-driven insights thatin turn generate impressive businessresults. Elizabeth Craig co-authored,with Peter Cheese and Robert J. Thomas,The Talent Powered Organization:Strategies for Globalization, TalentManagement and High Performance

(Kogan Page, 2007), one of the firstsystematic efforts to chart a strategyfor talent management in the globalenterprise.

In 2008, as a follow up to these books,Jeanne G. Harris and Elizabeth Craig,together with Henry Egan, launcheda new research project focused on theunique challenges of managing ana-lytical talent. We interviewed dozensof executives and surveyed more than1,367 full-time employees. A compre-hensive, web-based survey measuredthe personal engagement, work attitudesand career motivations of 799 analystsand 568 non-analysts. The respondentswere U.S. based employees of compa-nies with at least $50 million in annualrevenues. They represented a widevariety of industries and worked insuch functional areas as finance,IT, operations and production, R&D,marketing and sales.

Our survey sought to determinefirst, what engages analytical talent,and second, what influences analysts’intentions to stay with their companies.23

We assessed engagement by averagingtheir responses to nine questions thatasked participants to describe thefrequency with which they displayengagement behaviors at work. Weassessed executives’ intentions to stayby averaging their responses to threequestions that asked about their desireto and expectations and likelihood ofstaying with their firms now and inthe future. We conducted a series ofstatistical analyses on the survey datato determine what it takes to engageand retain analytical talent. We alsointerviewed dozens of executives tofind out how companies are managingthese challenges.



About the Research

To investigate the relationshipsbetween management practices andanalysts’ engagement and intentionsto stay, we conducted two types ofstatistical analyses. Multiple regres-sion analysis was used to determinethe influence of dozens of aspects ofemployees’ experience at work (e.g., joband career opportunities, organizationculture, management practices, etc.)on analyst engagement and intentionsto stay. We compared the significantpredictors for analytical talent to thosefor non-analytical talent to determinethe factors that are uniquely importantto analysts.

In this report of our findings, wedescribe the four variables thatemerged from the analyses as statisti-cally significant predictors of analystengagement or intentions to stay.To provide simple illustrations of theeffects of the four significant predictors,we report our findings from post-hoccross-tabulation analyses of businessinsight, role clarity, technical skills,supportive management, engagementand intentions to stay.

Acknowledgement: Interview findingscontributed by nGenera Insight.





Appendix: About the Sample

Our sample compared analytical talent (n=799) with non-analytical talent (n=568).

Type of talent Type of analytical talent

42%

n=1367 n=799

Non-analyticaltalent

Analyticaltalent

Analyticalchampion

Analyticalprofessional

Analyticalsemi-professional

Analyticalamateur

58%

52%

28%

11%9%

Respondents came from many industries.

Industry in which you work

0 15105 20 25 30

Analytical talent Non-analytical talent

Financial services

Communications and high tech

Health and life sciences

Consumer goods and services

Retail

Resources

Transportation and travel services

Public service

Automotive and industrial equipment

Professional services

Other

24.9%12.5%

20.5%16.2%

8.3%9.2%

7.8%6.9%

7.8%20.2%

6.1%5.6%

6.1%8.3%

5.9%6.9%

4.4%5.8%

4.4%4.2%

3.9%4.2%

Two-thirds of individuals in the sample worked for large organizations with more than US$1 billion in revenues,and over half of the companies employed more than 25,000 people.

Revenues

Number of employees

0 10 20 30 40 50 60 70


$1bn or more

$100m-$999m

$50m-$99m

65.6%

59.9%

21.3%

22.4%

13%

17.8%

0 10 20 30 40 50 60


25,000 or more

10,000-24,999

1,000-9,999

Fewer than 1,000

Don’t know

22%

21.3%

12.5%

11.4%

53.4%

52.8%

10%

10.2%

2%

4%





Survey participants worked in various corporate functions.

Primary functional work area

IT

Operations/Production

Finance

Analytics

Sales

Customer service, support

Engineering

Administrative

R&D/Scientific

Marketing

Clerical, processing

Consulting

Research

Exective management

Human resources

Business development

Administering health and mental health services

Purchasing

Account management

Education

Legal

Creative, design

Distribution

Merchandising

0 2 4 6 8 10 12 14

13%9.5%

10.5%12.3%

9.9%3.2%

8.6%1.4%

8.4%12.1%

7%8.1%

5.6%6.3%

5.0%6.0%

3.6%1.9%

2.9%1.9%

2.6%3.7%

2.3%1.8%

1.8%0.9%

1.6%1.8%

1.6%4.2%

1.5%1.2%

1.3%1.6%

1.3%0.9%

1%0.7%

1%2.1%

1%0.9%

0.9%0.4%

0.9%1.8%

0.3%1.1%




The sample covered all levels within the organization, from junior staff to chief executives.

Role in the organization

0 10 20 30 40 50 60 70 80


CEO

Senior manager

Manager

Individual contributor

0.9%

0.2%

28.5%

24.6%

60.7%

70.2%

9.9%

4.9%

About the AuthorsJeanne G. Harris ([email protected]) is an executive researchfellow and director of research at theAccenture Institute for High Performance.With Thomas H. Davenport, she is theauthor of Competing on Analytics:The New Science of Winning (Boston:Harvard Business School Press, 2007)and, with Davenport and Robert Morison,of Analytics at Work: Smarter Decisions,Better Results (Harvard Business Press,February 2010). She is based in Chicago.

Elizabeth Craig ([email protected]) is a research fellowat the Accenture Institute for HighPerformance in Boston. She is theauthor, with Peter Cheese and RobertJ. Thomas, of The Talent PoweredOrganization: Strategies forGlobalization, Talent Managementand High Performance (New York:Kogan Page, 2007).

Henry Egan ([email protected]) is a senior specialistwith the Accenture Institute forHigh Performance in London.



Notes1 “Competing Through Business Analytics

to Achieve High Performance,” Accenture

Information Management Services,

December 2008.

2 Thomas H. Davenport and Jeanne G. Harris,

Competing on Analytics: The New Science

of Winning, (Boston: Harvard Business

School Press, 2007).

3 Thomas H. Davenport, Jeanne G. Harris and

Robert Morison, Analytics at Work: Smarter

Decisions, Better Results (Harvard Business

Press, 2010).

4 Peter Cheese, Robert J. Thomas and Elizabeth

Craig, The Talent Powered Organization:

Strategies for Globalization, Talent Management

and High Performance (Kogan-Page, 2007).

5 When Accenture surveyed 254 executives

in 2008, nearly three-quarters said that

their companies were increasing their business

analytics usage. See “Competing Through

Business Analytics to Achieve High Performance,”

Accenture Information Management Services,

December 2008.

6 Thomas H. Davenport and Jeanne G. Harris,

Competing on Analytics: The New Science of

Winning, (Boston: Harvard Business School

Press, 2007).

7 Chris Donnelly,“Shopping for Scarce Retail

Talent: Driving High Performance in the Retail

Industry Through Strategic Talent Management”

(Accenture research report, 2009).

8 On September 21, 2009, Netflix declared

“Bell Kor’s Pragmatic Chaos,” a global group

of researchers, scientists, and engineers, as

winners of its $1 million (US) contest to

improve Cinematch. The winning entry improved

the model’s performance by 10.6%.

9 Raymond R. Panko, “What We Know

About Spreadsheet Errors”, Journal of End

User Computing 10 (1998): 15-21. Revised

May 2008: http://panko.shidler.hawaii.edu/

SSR/Mypapers/whatknow.htm.

10 We took into account other factors that are

known to affect employees’ engagement

and intentions to stay by controlling for age,

tenure, education, gender, level of responsibility

(dividing jobs into manager, manager of

managers, manager of senior managers, and

CEO) and company size.

11 Elizabeth Craig and Lauren Ready, How to

Create and Sustain a Highly Engaged Workforce,

Accenture Institute for High Performance

research report, forthcoming 2010.

12 According to one study, companies with highly

engaged workforces enjoy better performance—

up to a 103 percent higher success rate

(measured in terms of profits, productivity,

customer satisfaction and employee retention)—

than their counterparts with less-engaged

employees. See James K. Harter, Frank L. Schmidt

and Theodore L. Hayes, “Business-unit-level

relationship between employee satisfaction,

employee engagement, and business outcomes:

A meta-analysis,” Journal of Applied

Psychology, 2002.

13 J.R. Hackman and G.R. Oldham, “Motivation

Through the Design of Work: Test of a

Theory,” Organizational Behavior and Human

Performance, 1976.

14 In other research, we’ve found that engagement

levels contribute only a small amount

toward explaining why people stay or go.

See Elizabeth Craig and Lauren Ready, How to

Create and Sustain a Highly Engaged Workforce,

Accenture Institute for High Performance

research report, forthcoming 2010.

15 Percentages include responses of “likely” or

“very likely.”

16 Jeanne G. Harris, “How to Fill the Analytics

Talent Gap?” Strategy & Leadership, 2008.

17 The results of the cross tabulations presented

in this report represent comparisons between

the top and bottom quartiles for the relevant

variables. For example, in this case, more

than half of the analysts in the top quartile

of “business insight” were in the top quartile

of “engagement”, while just 18 percent of the

analysts in the bottom quartile of “business

insight” were in the top quartile of “engagement.”



18 According to research by the late Michael Driver,

as cited in: Brian O’Reilly and Antony J. Michels,

“Reengineering the MBA,” Fortune,

January 24, 1994.

19 http://strategic-hcm.blogspot.com/2008/05/

hr-for-innovation-google.html.

20 Bala Iyer and Thomas Davenport, “Reverse

Engineering Google’s Innovation Machine,”

Harvard Business Review, April 2008.

21 http://money.cnn.com/magainzes/fortune/

bestcompanies/2008/; http://money.cnn.com/

magainzes/fortune/bestcompanies/2007/full_list/

22 Sarah Fletcher, “Google: Recruiting and

Developing Top Talent.” Accessed Aug 26, 2009:

http://www.hrzone.co.uk/item/164452; The

average for high tech companies is 14 percent—

see “Winning the war for talent in the high-tech

industry,” SAP Executive Insight 2008.

23 Employees’ intentions to stay are routinely

used as an indicator of likely turnover behavior.

Extensive research has shown intentions to

stay to be powerful predictors of actual retention,

especially among professionals and white-collar

workers. See Aaron Cohen and Nadine Hudecek,

“Organizational Commitment-Turnover

Relationship across Occupational Groups,”

Group & Organization Management, 1993; and

R. Steele and N. Ovalle, “A Review and Meta-

analysis of Research on the Relationship between

Behavioral Intentions and Employee Turnover,”

Journal of Applied Psychology, 1984.

Copyright © 2010 AccentureAll rights reserved.

Accenture, its logo, andHigh Performance Deliveredare trademarks of Accenture.

About Accenture

Accenture is a global managementconsulting, technology services andoutsourcing company. Committedto delivering innovation, Accenturecollaborates with its clients to helpthem become high-performancebusinesses and governments. Withdeep industry and business processexpertise, broad global resources anda proven track record, Accenture canmobilize the right people, skills andtechnologies to help clients improvetheir performance. With more than177,000 people in more than 120countries, the company generated netrevenues of US$21.58 billion for thefiscal year ended Aug. 31, 2009. Itshome page is www.accenture.com.

About the Accenture Institutefor High Performance

The Accenture Institute for HighPerformance creates strategic insightsinto key management issues throughoriginal research and analysis. Itsmanagement researchers combine world-class reputations with Accenture’sextensive consulting, technology andoutsourcing experience to conductinnovative research and analysis intohow organizations become and remainhigh-performance businesses.

Documents

Accenture counting on_analytical_talent