HitsA metric from the early,
foolish days of the Web.Count people instead.Page viewsMarginally better than hits. Unless you’re displayingad inventory, count people.VisitsIs this one person visiting a hundred times, or are ahundred people visiting once? Fail.UsersThis tells you nothing about what they did, why theystuck around, or if they left.Followers/friends/likesCount actions instead. Find out how many followerswill do your bidding.LoginsBut what are they actually doing when they login?Logins don’t tell you about actions and value.Vanity metrics are bad!
Moms are crazy!(in a good
way)Engagement solved!• Messages to one another were on average 50% longer.• 115% more likely to attach a picture to a post they wrote.• 110% more likely to engage in a threaded (i.e. deep) conversation.• Friends, once invited, were 50% more likely to become engaged users.• 180% more likely to click on Facebook news feed items.• 60% more likely to accept invitations to the app.
Correlation lets youpredict the futureCausality
lets youchange the future“I will have 420 engaged usersand 75 paying customers nextmonth.”“If I can make more ﬁrst-timevisitors stay on for 17 minutes Iwill increase sales in 90 days.”Find correlation Test causalityOptimize thecausal factorCausality is a superpower, because it letsyou change the future.
The SaaS CustomerLifecycleCustomer Acquisition Costpaid
direct search wominherentviralityVISITORFreemium/trial offerEnrollmentUserDisengaged UserCancelFreemiumchurnEngaged UserFree userdisengagementReactivateCancelTrial abandonment rateInvite OthersPaying CustomerReactivation ratePaidconversionFORMER USERSUser Lifetime ValueReactivateFORMER CUSTOMERSCustomer Lifetime ValueViral coefﬁcientViral rateResolutionSupport dataAccount Cancelled Billing Info Exp.Paid Churn RateTieringCapacity LimitUpsellingrate UpsellingDisengaged DissatisﬁedTrial Over
•Stage: Revenue / Scale•Model: SaaS
(Paid)•Recruitment marketing andassessment software•Switched business models frommonthly subscription to pay perjob postingDoes recurring revenuework for everyone?
10xrevenue increaseoff of 3x in
salesvolume“People don’t do subscriptions for haircuts, hamburgers,and hiring. You have to understand your customer, whothey are, how and why they buy, and how they valueyour product or service.” - Ben Baldwin, co-founderLots of money!
EMPATHYSTICKINESSGROWTHRATEVIRALITYREVENUESCALELean AnalyticsStagesI’ve found a real,
poorly-met need thata reachable market faces.I’ve ﬁgured out how to solve the problem ina way they will adopt and pay for.I’ve built the right product/features/functionality that keeps users around.The users and features fuel growthorganically and artiﬁcially.I’ve found a sustainable, scalable businesswith the right margins in a healthyecosystem.“Gates” needed tomove forward
20%60%20%2%of visitors created an account(acquisition
/ Empathy)of sign-ups returned in the 1st month(engagement / Stickiness)of sign-ups were active after 6 months(engagement / Stickiness)convert from free to paid(Virality & Revenue)Buffer charges early to provepeople want the problem solved
• Target < 4% paid
churn (hitting 2% latelyon a monthly basis)•Anything over 5% means they don’t havea business that will generate positivemargin returns: the bucket is too leakyThe OMTM: Paid Churn
• Can we acquire more
valuable customers?•What product features can increase engagement?• Can we improve customer support?•Was a marketing campaign successful?•Were customer complaints lowered?•Was a product upgrade valuable?If Paid Churn: Why & Next Steps:Paid Churn = “business health” indicator•Are the new customers not the right segment?• Did a marketing campaign fail?• Did a product upgrade fail somehow?• Is customer support falling apart?
Some interesting benchmarksGrowth5% / week
(revenue or activeusers)Churn2% / monthEngaged visitors30% monthly users10% daily usersTime on site17 minutesPage load time< 5 secondsCLV:CAC3:1Mobile ﬁle size< 50MBFree to paid2% of free users
Draw a new linePivot orgive
upTry againSuccess!Did we move theneedle?Measure theresultsMake changes inproductionDesign a testHypothesisWith data:ﬁnd acommonalityWithout data:make a goodguessFind a potentialimprovementDraw a linePick a KPIThe Lean Analytics Cycle