Practitioner pattern · Trent Hommeyer (Works at Helium, the paywall builder/experimentation platform and paywall agency demoed in the episode)
Run 8 or 9 paywall tests at once and keep the one that wins
Usually 8 are a wash and 1 makes a difference
Workedexperimentationpaywall
What they did
Use a remote paywall tool (no app update needed) to test many variants at once: three plans on one paywall, different default-selected plans, colour schemes, trial lengths (1, 3, 7 days). Customers run eight or nine experiments simultaneously.
What happened
Typically eight are a wash and one moves the metric; that becomes the new paywall.
“we have people running like eight or nine experiments at a time just to like see what's up. And like eight of them are a wash, one of them makes a difference, and that's their new paywall.”
“Try a one-day trial. Try a three-day trial, try a seven-day trial, just just just try it out.”
Related topics
Get tactics like this every Monday
New tactics from the week's founder interviews, each linked to where it was said.
More from this episode
- Put a Flappy Bird game in the paywall; higher scores unlock up to 50% offBeat the existing paywall in an A/B test on 10% of users; no numbers given
- Show the paywall half-empty, then slide the trial details in from the sideWins 'really aggressively' in Helium tests, per the guest; no numbers
- Keep small animations moving in the onboarding background to hold focusWith animations beat without in a Helium test; no numbers
- Add a 'creating your experience' screen near the end of onboardingConversions up 11%, a founder told the host
- In a sleep-sounds app, play the sounds during onboardingConverts better, confirmed by two competing apps; no numbers
- For older users, ask 'Who referred you?' with options like 'My personal trainer' or 'TV ad'Works well with older audiences, per the guest; no numbers
- Test separate paywalls for users over and under about 35
- Charge through Stripe with an in-app Apple Pay prompt, not a trip to SafariBetter trial-to-paid, retention and ARPU; first conversion slightly lower
- App Masters: Send buyers from the paywall to a web checkout page in SafariBig drop-off every time; did worse than Apple in-app purchase
- Build for iPad, where few competing apps bother to go$1.4M of a client's $2M ARR came from iPad
- Default to a 7-day trial; offer 14 or 30 days to people who close the paywallConverts better as a fallback, clients say; anecdotal
- Helium: Ask sign-ups for their revenue; call the big apps, let small ones straight in