Retention-Friction Diagnosis
Use strong retention and weak growth to locate an access problem
- Difficulty
- Moderate
- Time to result
- ~weeks to results
- Steps
- 6
- Confidence
- 99%
Early Strava had low community growth but strong engagement and subscription conversion among people who endured its setup. Users needed a third-party GPS device, a computer, a cable, and a file upload before experiencing the product. Horvath interpreted the combination as evidence that the experience had value but access was badly designed. A basic mobile tracking app removed the purchase and upload barriers, and he says sign-ups jumped from roughly 100 a week to 10,000 on launch day, with an App Store feature later producing 100,000 in a day. That experiment also exposed a second mistake: users expected the results and community experience on mobile too. The method is to separate acquisition from retention, remove the largest entry friction, observe the response, and then complete the experience on the form factor that unlocked adoption.
Origin
Strava launched as a web-based cycling service that depended on third-party GPS hardware and computer uploads. Strong post-entry engagement but slow growth led the team to test mobile tracking and then rebuild the complete experience for mobile.
Core principles
- 01Strong engagement after entry can coexist with poor community growth
- 02High conversion among a small base is evidence of value, not proof of accessible distribution
- 03Remove the largest entry cost before rebuilding everything
- 04Meet users on the device where the full experience belongs
- 05A growth breakthrough can still expose the next product mistake
How to run it
- 1
Split the funnel
Measure how many people enter separately from what they do after reaching the product's value.
Pro tip Keep acquisition, activation, retention, engagement, and paid conversion distinct.
Watch out A blended growth number can hide a strong core behind a weak entrance.
- 2
Test for core value
Look for persistent use, engagement, and willingness to pay among people who complete onboarding.
Pro tip Use several post-entry signals rather than a single enthusiastic cohort.
Watch out A highly selected niche may not represent the wider audience.
- 3
Map entry friction
List required purchases, devices, transfers, forms, delays, and context switches before first value.
Pro tip Experience the onboarding with no existing equipment or insider knowledge.
- 4
Remove the largest barrier
Build the smallest change that eliminates the most consequential cost or step without replacing the whole product.
Pro tip Strava first made the phone a tracking device people already owned.
Watch out Do not introduce a simpler entrance that points to a broken destination.
- 5
Measure the unlock
Compare entry and activation rates immediately after the change while accounting for external distribution such as featuring.
Pro tip Separate product effects from platform promotion where possible.
Watch out Horvath's launch figures do not prove every friction reduction will create comparable growth.
- 6
Follow the user to value
Observe what newly acquired users expect to do next and move the complete experience to the form factor that unlocked entry.
Pro tip Treat the first growth spike as another diagnostic event, not the end of product work.
In the wild
Early users had to buy a GPS device, connect it to a computer, transfer a file, and upload it to Strava. A mobile app let people begin with the phone already in their pocket, producing what Horvath describes as a dramatic acquisition increase.
→ The test supported the diagnosis that access friction, rather than absence of value among existing users, was suppressing growth.
Common mistakes
Calling low growth no demand
Strong retention and conversion after a difficult onboarding can indicate that entry, not the core experience, is the immediate problem.
Stopping at mobile capture
Strava initially expected mobile users to visit the website for results and had to rebuild the full experience on mobile.
Attributing all growth to product
Horvath says an App Store feature amplified adoption, so platform distribution must be separated from the access change.
Is it for you?
Best for
It is best for products with a small but engaged user base and a measurable, removable barrier before first value.
Not ideal for
It is not ideal when retention is weak, the product causes harm, or growth is constrained by a factor the proposed access change cannot address.
From the transcript
“we had a low community size, but a high conversion rate”
“We saw off-the-charts community growth in the first week”
“what can unlock the community growth is the form factor of reducing the frictions”
From the episode
Strava Founder: How I Motivated 100 Million People To Stay Active: Michael Horvath