Homogenize the Early Experience for Retention
Move retention by fixing unlucky bad first experiences, not by nudging people about to churn
- Difficulty
- Moderate
- Time to result
- ~months to results
- Steps
- 3
- Confidence
- 92%
Retention is the culmination of the whole product experience, so the highest-leverage way to move it is to inflect the early user experience — specifically by finding customers who had a bad first experience by luck rather than by fit, and pulling those below-average experiences up to average.
Origin
A pattern Dan Hockenmaier repeatedly observed across consumer marketplaces, illustrated with the Uber/Lyft driver first-week earnings guarantees.
Core principles
- 01Retention is the culmination of the entire product experience, so rarely go straight to the source with more email/push
- 02By 3-6-12 months a customer has already formed a strong opinion; the first experience is where you can still teach value
- 03Look for variability — customers having a bad experience who shouldn't be — and homogenize it
- 04Correlational 'best users do X' analyses rarely move users from bucket B to bucket A
How to run it
- 1
Focus on the earliest lifecycle
Concentrate on the first week or month, where you still have the opportunity to prove the product's value before opinions harden.
Watch out Do not try to move a retention metric by sending more email or push notifications straight to the user — work core product levers, not growth-product levers.
- 2
Find unlucky bad first experiences
Look for variability in the early experience — customers who had a bad time by luck of the draw (a cancellation, a low-density area) rather than because the product is wrong for them.
- 3
Homogenize toward the true average
Streamline or guarantee the experience so those below-average first experiences are pulled up to average, showing the customer what the product is really like.
Pro tip Guarantees (like a first-week earnings floor) are a way to prove to a participant what the long-term experience will be like.
In the wild
A new driver may, purely by luck, earn a bad hourly rate their first week — a customer cancelled, or they were in a low-density area. The driver doesn't know that isn't representative and may conclude they only make three dollars an hour and never return. Uber and Lyft dueled to guarantee the highest first-week or first-month earnings.
→ Pulling below-average first experiences up to average proves what the experience is really like and drives much better retention curves.
Common mistakes
Spinning up resurrection efforts too early
Churned users are the pool who tried the product and decided against it; it is very hard to convince them otherwise. Wait until you have exhausted earlier-funnel efforts before investing in resurrection.
Copying what best users do onto everyone
The 'our best users do X, so make others do X' pattern almost never works because something unique about those customers drives the behavior — correlation rarely moves bucket B into bucket A.
Is it for you?
Best for
Growth and product teams at consumer marketplaces or apps trying to move a stubborn retention metric.
Not ideal for
Situations where the core product genuinely doesn't fit the churned users, or where onboarding is already homogeneous.
From the transcript
“retention is a tough measure to work on because it is the culmination of the entire product experience”
“the biggest wins in retention come from inflecting the early user experience”
“one really valuable lens is to look for variability in that experience”
“you see both of those companies dueling it out to guarantee the highest first week”
“our best users do X so why can't we make other users do that same thing”
“typically you want to wait to spin up Resurrection efforts until you've exhausted some of these earlier”
From the episode
Developing a growth model + marketplace growth strategy
Dan Hockenmaier (Faire, Thumbtack, Reforge)