Unlike outsourcing or technology, analytics packaged with external intelligence can add exponential value to organisations and individuals. Maturity models can turn out to be a myths sold by traditional consultants or large tech firms to upsell :) upgrades
Delivering business value using analytics & technology without worrying about maturity models (2)
1. Deliveringbusinessvalue usinganalytics& technologywithout worrying about maturity models
By Gaurav Vasu
DeliveringValue Through Analytics
Organizationsacrossindustriesare inthe processof increasingdigital engagement,closely
integratedwith analytics,fortheircandidates,employeesandcustomers.Whenwe meet
organisationsattemptingtodobiggerandbetterthings,we oftenturnto the maturitymodel
concept,bothas a meansof diagnosingthe level of the firm’scapabilitiesinagivenarea, andof
explaininghowthatcapabilitycanbe enhancedtoevergreaterheights.
Maturity modelsdescribethe characteristicsof maturityacrossa multi-pointscale,fromzeroorone
(typicallyachaoticstate),throughtosay five,where the capabilityis atthe highestlevel of maturity
and createsa genuine pointof difference forthe firm.Foranalytics,the journeyhasmanylayersand
it pushesfromtransactional reportingof lagmetricstoproactive reportingof leadmeasuresto
predictionandprescriptive analytics/insightsinvariousbusinessscenarios.
{Circle denoteswhat%of organizationsare atwhat stage of analytics}
Source of picture:BersinbyDeloitte |DUPress.com
The Myth of Maturity Models
In the domainof advancedanalyticsandbigdata, large analyticsconsultingprovidersorcore
analyticfirmsuse a varietyof factorsto pin-pointafirmonthe analyticsmaturityscale:
2. Let usfirstbreak the mythsof traditional maturitymodelsandshare learningsfromnotsosuccessful
digital transformation&analyticsjourneysof midtolarge size organisations.
Let me share withyoua couple of examplesfrommyresearch,observationsandinteractions
amongstlarge IT servicesfirmsinIndia*.
Example 1:
A topmultinational ITservicesfirmwithlarge offshore presence tookaboldstepa few yearsagoto
automate descriptivebusinessandHRreportingcompletelythroughQlik,assumingitwouldbe able
to expeditedecisionmakingtopositivelyimpactbusiness&HR metrics.Inthe thirdyear of its
journey,full-scale executionof all businessandHRreportsis still onwithverylittle impacton
decisionmaking.Fullcredittothe organisationforreducingreportingTAT(Clickof buttonvs.
reactive) andreporting&analyticsteamoptimisationsavingdollars.However,thisisanotional
savinggiventhe investmentmade intovendorevaluationtime,leadershiptime &costand internal
teamdeploymenttime &cost whichwouldeasilytake another2to 3 yearsto recover.
Challenges/learnings fromthe longexecutioncycle:
Technologyintegrationcapabilities(Synergiesof buyer&vendorsystems)
Vendordomainknowledgeforfastertransition/automation
100% reliance onreportingautomationcanstall shorttermbenefitsof semi-automateddescriptive
reports
Example 2:
Anotherlarge multinational IT&consultingservicesfirmlookedtooptimisetalentacquisitioncost&
time throughadvancedalgorithmsandanalyticsthree yearsback.The firmstartedthe journeyby
automatingcandidate screeningprocessthroughNLP&artificial intelligence technologyplatform,
followedbyintegratingpredictive analyticstoscore candidatesonemployability/fitment,forecast
probabilityof offeraccepttoevencandidate performance/stickinessthroughBig Data.
The myth withwhichthe organisationwentaheadwasthe factthat beingaPCMM level 5company,
theirhiringprocesses,candidate data,jobdescriptions,hiringmanagerbehaviourandhistorical
Is the firmdealingwiththe basics
(reliableproductionof truthful
reportingof internally-sourceddata),
or is itin the realmof solving
complex questionsusingalgorithms
that don’tjustanalyse internal data,
but alsodata fromoutside the firm?
Dependingonwhere the firmsits,
whatwouldthe nextgenerationof
sophisticationlooklike?
Beinga numbers-drivenprofessional,
do I feel it’sworththe waitforthe
investmenttoreapbenefits?
Do organisationsespeciallylarge
oneswhoare PE funded,listedon
stock exchangesoreven
entrepreneur-driven,have patience
to waitfor the results?
QuestionstoAsk
3. employeedataforlike-to-likeprofileswasstructured,cleanandfullyreliable.The pilotandactual
launchshowcasedthatwhile candidate datawasrich,the internal datato derive keyinsightsfrom
predictive analyticswasextremelyweak.Asaresult,benefitswere reapedonlyonautoscreening.
Challengesandlearningsfromthe pilot&executionexperience:
Capabilities –Runningapilotto testpredictive analyticscapabilitiesonsmall scale
(Capabilitiesof notonlyvendorbutbuyerdataquality,dataenvironment&availability.Any
variables withlessthan30%data availabilitycannotbe usedforanypredictive or
prescriptive analytics)
The vendorfailedtoassessdescriptive reportinginbuyerorganisation(Organisations
withoutmatureddescriptive reporting/analyticscapabilitiesfinditdifficulttomove to
predictive analytics)
Cuttingedge AItechnologywithweakdatayieldslittle ornoimpact
It isnecessarytoissue a caveathere:The maturitymodel of youranalyticspractice needstomarry
the organisational maturitylevel.Keyquestionstoconsiderhere are:
Is there leadershipbuy-inforprescriptiveactionsfrompredictive analytics?
Are there sufficientprocessesinplace tocapture cleandata?
Doesorganizationhave acentral data warehouse?
Do we have technologytosupportadvanceddataanalytics?
Doesorganizationhave datarequiredtoaccuratelypredict?
Leadershipbuy-iniscomparativelyeasierinfinancial analyticswhere we dealwithnumbersand
predictnumbers;whereas,it’sanuphill taskwhenthe managementproblemrevolvesaround
humanbehaviourandthe attitude rangesfromcompletelypredictabletoabsolutely
undecipherable!
Also,itisimportantto understandthatStatisticsisonlyameansto an endwhichpreventsahelter-
skelterapproachtoservice delivery.Without doingpredictive modellingandcreatingcausation
frameworks,there’sstill muchvalue thatcanbe deliveredtobusiness.A leapinthe analyticsvalue
chainwithoutthe same leverinthe organizational maturitywouldrenderthe outcome useless.
Here are a fewexamplestodemonstratethe value deliveredatdifferentmaturitylevelswithoutthe
constrainsof stickingtogradual movementupanalyticsvalue chain:
AirBnBand Indianbed-linenindustry –Nearlya decade back,a handful of financial analysts
supportingHNI’sransimple descriptive&correlationanalytics(graphsanddashboards) ongrowth
of bed-linenindustryinIndiaandmergedwithcausationanalyticstoidentifypossible indirect
sourcesof demandsforlistedbed-linencompanies.Thissort of datacan be a goldmine fordecisions
aboutcommunitygrowth,productdevelopment,andresource prioritization.Thisledtoaconclusion
that growthinthe numberof roomslistedonAirBnBbyhome ownersandthe pushtowards
cheapergoodalternativesto the establishedhotel industrywouldpushthe demandforbed-linens
frommanufacturersinIndia.Five yearsdownthe line,bothlistedandunlistedplayershave seena
huge surge inboth topline andbottomline anddecade downthe line few playersbecome
multibaggers.
4. Financial AnalyticsforendHNIcustomer – The financial servicesindustryindevelopednationsare
trulyleveragingthe powerof prescriptiveanalytics,providingprescriptive actionstocustomerson
bankingandinvestmentdecisions,based onfactorssuchas personal health,financial status,
weatherconditionsincitiestheylivein,political orecological changesandmanymore suchdirect
deriveddatawithfinancial goal set.
Prescriptive AnalyticsandConsultingforGovernments –Mostmaturedcountrieswhohave
connectedsystems,highendtechnologyandcitizendatacapturedoveryearsare,infact, ina better
positiontoleverage prescriptive analyticsinpolicy,political andeconomyrelateddecisions.Virtual
Singapore isone of the bestexampleswherepolicymakers,statesmenandcitizenscanbuildortest
decisionsthroughuse of IoTand prescriptiveanalytics.The USgovernmenthasa teamof data
scientiststestingvariousdecisionstooutcome of variouspoliciesimplementedaround the worldto
helpenhance the decisionmakingprocess.Thisisaclassicexample of startingwithdescriptive
reportingtodescriptive insightsgenerationtoslowlymovetowardspredictivewhichisextremely
mature & requireshuge investments.
How to Derive Value?
Value comesfromthingsthatmatterat thatpointin time andfor the overall organisational purpose.
Almost80% of the reportingdone inmostorganizationsistransactional withmetricsthatare of little
value —that’swhere we shouldstart.
The right metricsprovide acontextaroundwhichperformance canbe analysed.Withthe wrong
metrics,anexecutive summarygetsreducedtoananecdotal commentary.One of the classic
examplesof businessmetricinHR:Employee AttritionReporting,PredictiontoPrescription.
Reportingemployeeattritionbyvariousbands/locations/skillsetc.renderslittle ornovalue unless
keyemployeeclusters/groupstobe retainedare identified,reported,predictedandprescriptive
actionstakenon thisclustertoimpact employee attritionasa metric.
Such outcomespresentedinacompellingfashionwinoverthe management’strustandgetthemto
rethinkprocessesandworkflows,investintechnologyandinfrastructure,andadapttothe
organisational changesthatarise outof youranalytical frameworks. And when theorganisational
and businessneedsarein synergy,you moveup theanalyticalvaluechain,leaving themanagement
to answerjustonequestion,Arewe awareof the next bigthing?
Analyticsteamneedstodeliverydescriptive &predictive modelswith Intelligence, Insightsand
Prescriptive actionsto drive value andnot justgoodlookingdashboardsorstatistical models.Thisis
where a combinationof statisticians,datascientistsanddomainexpertscantogether addvalue to
answerthe questionof nextbigthing?
Important for Analyticsteams to focus on Resultsbut imperative for leadershipto be ready to
convert prescriptive insightsto actions
Significantuntappedvalue liesindatathatalreadyexistsinmost organizations,andanalyticsteam
needstoassessrequiredcapabilitiesthatcaneffectivelyexploitthisdata.Howeverbothleadership
teamand analyticsteamneedstoestablishwithaclearfocusontangible businessvalue.
Basedon researchone of the keyreasonsforfailure of internal analyticsteaminlarge organizations
across varioussectorsstartingfromPharma,Banking,Financial ServicestoIThas beenlackof
synergiesandunderlyingintentof settingupanalyticsteambetweenleadership&analyticsteam.
Analyticsteamneedsto establish anAgile Analyticsdataarchitecture andmethodologytoaddresses
ever-changingbusinessrequirementsandopportunitiesinaway that can evolve alongwiththe
businesstobecome asource of genuine strategic value (LinkitbacktoBalance Scorecard or
organizationgoals).Businessusersandleadershipneedstobe preparedtoact on insightsfrom
descriptive topredictive analyticmodelstosee true value of investmentsmade.
5. About the Author: Gaurav Vasu
Gaurav Vasuis Global HR Market Intelligence &Analyticsleadata leadingglobal ITservices
company.He hasworkedwithCEOoffice,CHRO’sandseniorHR leadershiptoshape the business
strategy,identifyhumancapital implicationsanddesignpeoplepracticestoenhance performance
and productivity.He isone of the top Industryexpertsinthe Research,ConsultingandAnalytics
domain.
He alsospecializesinGrowthConsulting(IT/ITES),MarketEntryStrategy,IndustryAnalysis&
Assessment(India,China&Philippines),TalentSupplyMapping,VendorAnalysis,PeerGroup
Benchmarking,FinancialAnalysis(DiscountedCashFlows,Relative Valuations,Simulation,etc.) and
Wargaming.
Duringhis12+ yearsof experience,Gauravhashelpeddeliveredconsultingbygrowingthe research
and analyticsvalue chainincompaniessuchasHCL, Accenture,Zinnov,andKnowledge Faber&
Nirvana.