SaaS Metrics – A Guide to Measurin and Improvin What Matters This is a guest post by David Skok. David is a five time serial entrepreneur S U B S T O T B L O G First Name * Email * Submit Analytics Blogging Branding Case Studies Content Marketing Conversion Demand Generation Email Marketing B R O W M A R T O P I
183
Embed
The SaaS Metrics – Measuring and Improving What Matters
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
This articleis acomprehensiveanddetailedlook at thekeymetricsthat areneeded tounderstandandoptimize aSaaSbusiness. Itis acompletelyupdatedrewrite ofan olderpost. Forthis
businesses.Traditionalbusinessmetricstotally failto capturethe keyfactorsthat driveSaaSperformance.In the SaaSworld,there are afew keyvariablesthat makea bigdifferenceto futureresults.This post isaimed athelpingSaaSexecutivesunderstandwhichvariablesreallymatter,and how tomeasurethem andact on theresults.
differentbecausetherevenuefor theservicecomesover anextendedperiod oftime (thecustomerlifetime). Ifa customeris happywith theservice,they willstickaround fora longtime, andthe profitthat can bemade fromthatcustomerwillincreaseconsiderably.On theother handif acustomerisunhappy,they will
they willchurnquickly,and thebusinesswill likelylosemoney ontheinvestmentthat theymade toacquirethatcustomer.Thiscreates afundamentallydifferentdynamic toatraditionalsoftwarebusiness:there arenow twosales thathave to beaccomplished:
1. Acquiringthecustomer
2. Keepingthecustomer(tomaximize
maximizethelifetimevalue).
Because oftheimportanceofcustomerretention,we will seea lot offocus onmetricsthat helpusunderstandretentionand churn.But firstlet’s look atmetricsthat helpyouunderstandif yourSaaSbusiness isfinanciallyviable.
TheSaaS
TheSaaSP&L /CashFlowTroughSaaSbusinessesfacesignificantlosses inthe earlyyears (andoften anassociatedcash flowproblem).This isbecausethey haveto investheavilyupfront toacquire thecustomer,butrecoverthe profitsfrom thatinvestmentover a longperiod oftime. Thefaster thebusinessdecides to
SaaS isusually a“winner-takes-all”game, andit isthereforeimportantto grabmarketshare asfast aspossible tomake sureyou are the
you are thewinner inyourspace.Providedyou cantell a storythat showsthateventuallythatgrowth willlead toprofitability,WallStreet,acquiringcompanies,andventureinvestorsall rewardhighergrowthwith highervaluations.There’salso apremiumfor themarketleader in aparticularspace.
this topichere:StartupKiller: TheCost ofCustomerAcquisition,and here:How SalesComplexityimpactsCAC.)
IsyourSaaSbusinessviable?In the firstversion ofthis article,Iintroducedtwoguidelinesthat couldbe used tojudgequicklywhetheryour SaaSbusiness isviable. Thefirst is agood way
good wayto figureout if youwill beprofitablein the longrun, andthe secondis aboutmeasuringthe time toprofitability(which alsogreatlyimpactscapitalefficiency).
Over thelast twoyears, Ihave hadthe chanceto validatetheseguidelineswith manySaaSbusinesses,and itturns outthat theseearlyguesseshave held
have heldup well.The bestSaaSbusinesseshave a LTVto CACratio thatis higherthan 3,sometimesas high as7 or 8. Andmany ofthe bestSaaSbusinessesare able torecovertheir CACin 5-7months.HowevermanyhealthySaaSbusinessesdon’t meettheguidelinesin the earlydays, butcan seehow theycanimprovethe
businessover timeto getthere.
Thesecondguideline(Months toRecoverCAC) is allabout timetoprofitabilityand cashflow.Largerbusinesses,such aswirelesscarriersand creditcardcompanies,can affordto have alongertime torecoverCAC, asthey haveaccess totons ofcheapcapital.Startups,on the
on theotherhand,typicallyfind thatcapital isexpensivein the earlydays.Howevereven ifcapital ischeap, itturns outthatMonths torecoverCAC is avery goodpredictorof how wella SaaSbusinesswillperform.Take alook at thegraphbelow,whichcomesfrom thesamemodelusedearlier. Itshows howthe
theprofitabilityis anemic ifthe time torecoverCACextendsbeyond 12months.
I shouldstress thatthese areonlyguidelines,there arealwayssituationswhere itmakessense tobreakthem.
ThreeusesfortheSaaSGuidelines
Guidelines1. One
of thekeyjobsof theCEO istodecidewhento hittheacceleratorpedal.Thevalueofthesetwoguidelinesis thattheyhelpyouunderstandwhenyouhave aSaaSbusinessthat isingoodshape,whereitmakes
makessenseto hittheacceleratorpedal.Alternativelyif yourbusinessdoesn’tmeettheguidelines,it is agoodindicatorthatthereismoretweakingneededto fixthebusinessbeforeyoushouldexpand.
Thosewithprimarilyannualcontracts,withsomecontractsformultipleyears.Heretheprimaryfocus ison ARR(AnnualRecurringRevenue),andACV(AnnualContractValue).
Most ofthe time inthis article,I will refertoMRR/ACV.Thismeans useMRR if youare thefirst kindofbusiness,or ACV ifyou are thesecondkind ofbusiness.Thedashboardshownbelowassumesmonthlycontracts(MRR).Howeverin thedownloadablespreadsheet,there is atab thatshows thesamedashboardfor thesecond
problembecomesdifferent.Imaginethat youhavebecomereally big,and nowhave amillioncustomers.3% churnmeansthat youare losing30,000customerseverymonth!That turnsout to be amuchhardernumber toreplace.CompanieslikeConstantContacthave runinto thisproblem,and it hasmade itvery hardfor them
In the topgraph, weare losing3% of ourrevenueeverymonth,and youcan seethat with aconstantbookingsrate of $6kper month,therevenuereaches$140kafter 40months,andgrowth is
Nowimaginethat welose 10 ofthem. OurCustomerchurn rateis 10%. Butif out ofthe tenchurnedcustomers,9 of themwere smallaccounts,and onlyone was alargeaccount.We wouldonly havelost $1,900in MRR.Thatrepresentsonly 3.4%RevenueChurn. Soyou cansee thatthe twonumberscan bequitedifferent.But each isimportant
importanttounderstandif we wanta completepicture ofwhat isgoing on inthebusiness.
GettingpaidinadvanceGettingpaid inadvance isreallysmart ideaif you cando itwithoutimpactingbookings,as it canprovidethe cashflow thatyou needto coverthe cashproblemthat wedescribedearlier in
earlier inthe article.It is oftenworthprovidinggoodfinancialincentivesin the formofdiscountstoencouragethisbehavior.The metricthat weuse totrack howwell yoursales forceis doing inthis area isMonths upFront.
Gettingpaid moreupfrontusuallyalso helpslowerchurn. Thishappensbecausethecustomerhas made
has madea greatercommitmentto yourservice,and ismore likelyto spendthe timegetting itup andrunning.You alsohave moretime toovercomeissues thatmight arisewith theimplementationin the earlydays.CalculatingLTV andCAC
The Metric“Monthsup Front”has beenused atbothHubSpotandNetSuite inthe past asa way to
Are welosingmost ofthecustomersin thefirstcoupleofmonths?
DoesChurnstabilizeaftersome
someperiodof time?
Then iftook someactions totry to fixchurn inearlymonths,(i.e withbetterproductfeatures,easier on-boarding,bettertraining,etc.) wewouldwant toknow ifthosechangeshad beensuccessful.The cohortanalysisallows usto do thisbycomparinghow morerecentcohorts(e.g. July in
(e.g. July inthe tableabove)comparedagainstJanuary.The tableaboveshows thatwe made abigimprovementin the firstmonthchurngoing from15% to 4%.
Twowaysto runCohortAnalysisThere aretwo waysto runCohortAnalysis:the firstlooks atthenumber ofcustomers,and thesecondlooks at
exampleillustrateshow thegraphwould lookif there isvery strongnegativechurn. Asyou cansee, theincrease inrevenuefrom thecustomersthat arestill usingthe serviceis easilyoutpacingthe lostrevenuefromchurnedcustomers.It is prettyrare to seethings lookthis good,but it is theidealsituationthat we arelooking for.For thosewonderingif this can
if this canbeachieved,onecompanyin ourportfolio,Zendesk,that hasnumbersthat areevenbetter thanthoseshown intheexamplebelow.
In thesituationabove, youwill need amorecomplexformula tocalculateLTV, as thevalue oftheaveragecustomerisincreasing
Similarly ifyou soldyour SaaSproduct toa 100persondepartment,and only10 peoplewere usingit, youwouldscore thatdifferentlyto 90peopleusing it. Sotherecommendationis that you
is that youcreate aCustomerEngagementScore,based onallocatingpoints fortheparticularfeaturesused.Allocatemorepoints forthefeaturesyou believeare moststicky.(Later onyou can goback andlook at thecustomerswhoactuallychurned,andvalidatethat youpicked therightfeatures asa predictorof whowould
satisfaction.Therecommendedway tomeasurecustomerhappinessis to useNetPromoterScore(NPS). Thebeauty ofNPS is thatit is astandardizednumber,so you cancompareyourcompanyto others.For moredetails onNetPromoterScore, clickhere.
Yoursalesforcemayhaveoversoldtheproduct,or soldit to acustomerthat isnot wellsuitedto getthebenefits
Youmay besellingtoSMB’s
wherea lot ofthemgo outofbusiness.It isn’tenoughthatwhatyou’resellingissticky.Whoyou’resellingit tomustalso besticky.
You arenotusing apricingschemethathelpsdriveexpansionbookings
The bestway to findout whycustomers
arechurning isto get onthe phonewith themand askthem. Ifchurn is asignificantpart ofyourbusiness,werecommendthat thefoundersthemselvesmakethese calls.They needto hearfirst handwhat theproblem is,as this is soimportantfor thesuccess ofthebusiness.And theyare likelyto be thebestpeople todesign afix for the
fix for theproblem.
TheImportanceofCustomerSegmentationIn all SaaSbusinessesthere willlikely comea momentwhere theyrealize thatnot allcustomersarecreatedequal. Asanexample,biggercustomersare harderto sell to,but usuallyplacebiggerorders,and churnlessfrequently.We need away to
way tounderstandwhich ofthese aremostprofitable,and thisrequiresus tosegmentthecustomerbase intodifferenttypes, andcomputethe uniteconomicsmetrics foreachsegmentseparately.Commonsegmentsare thingssize of ofcustomer,verticalindustry,etc.
Themetricsthat matterfor eachsalesfunnel,vary fromonecompanyto the nextdependingon thestepsinvolved inthe funnel.Howeverthere is acommonway tomeasureeach step,and theoverallfunnel,regardlessof yoursalesprocess.Thatinvolvesmeasuringtwo thingsfor eachstep: thenumber ofleads thatwent intothe top of
the top ofthat step,and theconversionrate to thenext stepin thefunnel (seebelow).
In thediagramabove,(mirroredin thedashboard),we show averysimplethreephasesalesprocess,withvisitorscoming toa web site,and someportion ofthemsigning upfor a trial.Then someof the trials
of the trialsconvert topurchases.As you cansee in thedashboard,we willwant totrack thenumber ofvisitors,trials andcloseddeals. Ourgoalshould beto increasethosenumbersover time.And we willalso wantto track theconversionrates, withthe goal ofimprovingthose overtime.
UsingFunnelMetricsinForwardPlanning
PlanningAnotherkey valueof havingtheseconversionrates is theability tounderstandtheimplicationsof futureforecasts.Forexample,lets sayyourcompanywants todo $4m inthe nextquarter.You canworkbackwardsto figureout howmanydemos/trialsthatmeans,and giventhe salesproductivitynumbers –how many
salespeoplearerequired,and goingback astageearlier,how manyleads aregoing to berequired.These arecrucialplanningnumbersthat canchangestaffinglevels,marketingprogramspendlevels, etc.
SalesCapacityIn manySaaSbusinesses,sales repsplay a keyrole inclosingdeals. Inthose
WhencomputingSalesCapacity, ifa newerrep is stillrampingand onlyexpectedto deliver50% ofquota,they canbe countedas half of aproductiverep. That isoftenreferred toas FullTimeEquivalentor FTE forshort.
addingsales reps,make sureyou alsohave aclear planof how youwill drivetheadditionalleadsrequired.
There ismuchmore thatcould besaid onthis topic,but since itis all verysimilar tomanaginga salesforce in atraditionalsoftwarecompany,we willleave thatfor otherblog posts.
andproduceless leadsfor moredollarsinvested.As a result,SaaScompanieswill needto beconstantlyevaluatingnew leadsourcesthat theycan layerin on topof the oldto keepgrowing.
Since theconversionrates andcosts perlead varyquiteconsiderably,it isimportantto alsomeasurethe overall
the overallROI byleadsource.
Growingleads fastenough tofeed thefront endof thefunnel isone of theperennialchallengesfor anySaaScompany,and islikely to beone of thegreatestlimitingfactors togrowth. Ifyou arefacing thatsituation,the mostpowerfuladvice wecan giveyou is tostartinvestingin InboundMarketingtechniques
techniques(see GetFoundusingInboundMarketing).This willtake timeto rampup, but ifyou can doit well, willlead to farlower leadcosts, andgreaterscalingthan otherpaidtechniques.Additionallythe typicalSaaS buyeris clearlyweb-savvy,andthereforevery likelytoembraceinboundmarketingcontentandtouchlesssellingtechniques.
WhatLeversareavailabletodriveGrowthSaaSbusinessesare morenumericallydriventhan mostotherkinds ofbusiness.Making asmalltweak to anumberlike thechurn ratecan have avery bigimpact onthe overallhealth ofthebusiness.Because ofthis wefrequentlysee a
see a“quant”(i.e. anumbersoriented,spreadsheetmodeling,type ofperson) asa valuablehire in aSaaSbusiness.AtHubSpot,BradCoffeyplayed thatrole, andhe wasable to runthe modelstodeterminewhichgrowthplaysmade themostsense.
Understandingthese SaaSmetrics isa key steptowardsseeing howyou can
you candrive yourbusinessgoingforward.Let’s lookat some ofthe leversthat theseimply asgrowthdrivers foryourbusiness:
David’swork as avalueaddedinvestor isbestknown for
known forhelpingJBoss,Netezza,AppIQ,CloudSwitch,Tabblo andDiligentTechnologiestosuccessfulexits.Davidserves onthe boardsofCloudBees,Digium(makers ofthepopularAsteriskOpenSourcePBX/telephonysoftware),Enservio,GrabCad,HubSpot,OpenSpan,andVideoIQ.He is alsoan investorin Meteor.