Customer Lifetime Value - Customer Analytics With R

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 7

8/2/2016

CustomerLifetimeValue|CustomerAnalyticswithR

CustomerLifetimeValue
Thebusinessproblem
Problem:wedon'tunderstandourcustomer.Wedon'tknowifaspecificcustomerisrelevantforourorganization.
Goals:
Wewanttoknowhowmuchvalueacustomerisgeneratingfortheorganization.
Wewanttoknowhowthegeneratedvalueisgoingtoevolve.
Wewanttosegmentcustomersbasedonvalue.
Why?Toperformspecificactionsfordifferentcustomergroups

Techniquetosolvethebusinessproblem
Wehavemanyoptionstounderstandcustomer'svalue.OneofthemostrelevantisCLV.CLVistheacronymofCustomerLifetime
Value.Weneedadefinition:
CLVisapredictionofallthevalueabusinesswillderivefromtheirentirerelationshipwithacustomer.
Thereisatoughpartinthisdefinition:howtoestimatefuturecustomerinteractions.
ThecalculationofCLVcanbebasedon:
ARPU/ARPA(HistoricCLV)
RFM(onlyapplicabletothenextperiod)
CLVFormula(historicCLV)
Probability/Econometrics/PersistenceModels/MachineLearning/Growthanddisseminationmodelssuchas:
MovingAverages,Regressions,BayesianInference,Pareto/NBD(NegativeBinomialDistribution)
https://josepcurtodiaz.gitbooks.io/customeranalyticswithr/content/chapter3.html

1/7

8/2/2016

CustomerLifetimeValue|CustomerAnalyticswithR

ThevaluegeneratedbyallourcustomersiscalledCustomerEquityandisusedtoevaluatecompanies(withnosignificantincome
yet).

MainConcepts
Therearemanyfactorsthataffectcustomer'svalueandmustbeconsideredintheCLVcalculation:
Cashflow:thenetpresentvalueoftheincomegeneratedbythecustomertotheorganizationthroughouttheirrelationshipwith
theorganization
Lifecycle:thedurationoftherelationshipwiththeorganization
Maintenancecosts:associatedcoststoensurethattheflowofrevenuepercustomerisachieved
Riskcosts:riskassociatedwithaclient
Acquisitioncosts:costsandeffortrequiredtoacquireanewcustomer.
Retentioncosts:costsandeffortrequiredtoretainanewcustomer.
Recommendationvalue:impactoftherecommendationsofacustomerinitssphereofinfluenceoncompanyrevenue
Segmentationimprovements:valueofcustomerinformationinimprovingcustomersegmentationmodels

CLVformula
Itis:
$$CLV=\sum^{T}{n=0}\frac{(p{t}c_{t})r_t}{(1+i)^t}AC
$$
where
$$p_t=$$pricepaidbythecustomerintime$$t$$,
$$c_t$$=directcostsforcustomerserviceintime$$t$$,
https://josepcurtodiaz.gitbooks.io/customeranalyticswithr/content/chapter3.html

2/7

8/2/2016

CustomerLifetimeValue|CustomerAnalyticswithR

$$i$$=discountrateorcostofmoneyforthefirm,
$$r_t$$=probabilitythattheclientreturnstobuyorisaliveintime$$t$$,
$$AC$$=acquisitioncost,and
$$T$$=timehorizontoestimate$$CLV$$.

Asimplification
Let'sconsiderthat$$p$$y$$c$$areconstantvaluesand$$r$$isadecresingfucntionintime,thentheformulabecomes:
$$CLV=\sum^{\infty}_{n=0}\frac{(pc)r^t}{(1+i)^t}=\frac{(pc)r}{1+ir}
$$

Finalconsideration
Generallywewillhaveadatasetforeachoftheconceptsoftheformula
We'llhavetoestimatetherest
Therefore,itisimportanttorememberthat:
Ourabilitytopredictthefutureislimitedbythefactthattosomeextentiscontainedinthepast.
Whatmathematicallymeansweareundersomecontinuityconditions(orthehyphotesisaretrue)

ImplementationProcess
[BU]DiscusswhetherCLVfitsasametricinourbusiness
[DP]Identificationandunderstandingofsourcesandmetadata
https://josepcurtodiaz.gitbooks.io/customeranalyticswithr/content/chapter3.html

3/7

8/2/2016

CustomerLifetimeValue|CustomerAnalyticswithR

[DP]Extract,transform,cleanandloaddata
[M]ChooseCLVmethod
[M/E]Analyzeresultsandadjustparameters
[D]Presentandexplaintheresults

Benefits
Thistechniqueprovidesthefollowingbenefits:
Abilitytocreatebusinessobjectives
Betterunderstandingofcustomers
Havingacommonnumericalanalysiscriteria
Abilitytohaveanalertsystem
Improvedmanagementofthesalesforce

Usecases
CreatemarketstrategiesbasedonCLV
CustomersegmentationbasedonCLV
Forecastingandcustomerevolutionpersegment
Createdifferentcommunication,servicesandloyaltyprogramsbasedonCLV
Awake"nonactive"customers
Estimatedthevalueofacompany(startup,inthecontextofacquisition)

HowtoimplementCLVusingR
https://josepcurtodiaz.gitbooks.io/customeranalyticswithr/content/chapter3.html

4/7

8/2/2016

CustomerLifetimeValue|CustomerAnalyticswithR

ARPU/ARPAasCLVaproximation
AverageRevenueperUser(ARPU)andAverageRevenueperAccount(ARPA)canbeusedtocalculatehistoricalCLV.Process:
calculatetheaveragerevenuepercustomerpermonth(totalrevenuenumberofmonthssincethecustomerjoined)
addthemup
andthenmultiplyby12or24togetaoneortwoyearCLV.
SupposeJosepandLauraareyouronlycustomersandtheirpurchaseslooklikethis:
CustomerName

PurchaseDate

Amount

Josep

January1,2015

$150

Josep

May15,2015

$50

Josep

June15,2015

$100

Laura

May1,2015

$45

Laura

June15,2015

$75

Laura

June30,2015

$100

SupposetodayisJuly1,2015.AveragemonthlyrevenuefromJosepis
$$(150+50+100)/6=50$$
andaveragemonthlyrevenuefromLaurais
$$(45+75+100)/2=110$$.
Addingthesetwonumbersgivesyouanaveragemonthlyrevenuepercustomerof$160/2=$80.Tofinda12monthor24month
CLV,multiplythatnumberby12or24.
https://josepcurtodiaz.gitbooks.io/customeranalyticswithr/content/chapter3.html

5/7

8/2/2016

CustomerLifetimeValue|CustomerAnalyticswithR

ThebenefitofanARPUapproachisthatitissimpletocalculate,butitdoesnottakeintoaccountchangesinyourcustomers'
behaviors

Let'sunderstandtheCLVformula
#Installpackages
install.packages("readxl")
install.packages("ggplot2")

#Loadpackages
library(ggplot2)
library(readxl)

#Loaddataintoadataframe
df

References
ModelingCustomerLifetimeValue
CLTV,Harvard
The16businessbenefitsofCustomerLifetimeValue
DominiqueHanssens'sResearch
BTYDpackage
KinshukJerath,PeterS.Fader,BruceG.S.Hardie,Customerbaseanalysisusingrepeatedcrosssectionalsummary(RCSS)
data,EuropeanJournalofOperationalResearch,Volume249,Issue1,16February2016,Pages340350,ISSN03772217,
http://dx.doi.org/10.1016/j.ejor.2015.09.002.
https://josepcurtodiaz.gitbooks.io/customeranalyticswithr/content/chapter3.html

6/7

8/2/2016

CustomerLifetimeValue|CustomerAnalyticswithR

Canthenegativebinomialdistributionpredictindustrialpurchases?JohnWWilkinson,GiangTrinh,RichardLee,NeilBrown.
JournalofBusiness&IndustrialMarketing201631:4
RajkumarVenkatesan,V.Kumar(2004)ACustomerLifetimeValueFrameworkforCustomerSelectionandResourceAllocation
Strategy.JournalofMarketing:October2004,Vol.68,No.4,pp.106125

https://josepcurtodiaz.gitbooks.io/customeranalyticswithr/content/chapter3.html

7/7

You might also like