Celebs · B2C · celebrity lookalike app
Drop the paywall after onboarding; show the lookalike result fast and ask later
More revenue per user; later paywalls converted the users who used to drop
Workedpaywallonboarding
What they did
A/B tested removing the end-of-onboarding paywall so users reach the 'moment of maximal intrigue' (their lookalike result) as fast as possible. Works because the app already had ~10 optimized conversion placements downstream (on seeing the result, on sharing, on reopening). Advice: keep the onboarding paywall (hard or trial) unless there is a clearly better moment later in the app to ask.
What happened
Big win in revenue per user; users previously lost at the onboarding paywall converted well further down the funnel. Earlier attempts at the same test had failed.
In their words
The Superwall Podcast · Patrick Stuart-Constant & Pablo Sánchez - $16M AI CEOs: How to Make $1.3M/Month With AI Apps“Removing a paywall at onboarding, funny enough, was a big win on Celebs. Generally, it's a good practice to have a paywall at the end of your onboarding. Here, the big win was removing it and actually just getting the users as quickly as possible to the moment of maximal intrigue, basically.”
“We can get conversion in 10 different places in the app. Probably it's a matter of we manage to optimize all down the final places, so it was making more sense to let the users go inside the app without the friction of a paywall, because our other placements are performing better.”
“So in the end, when you get your result, when you want to share, when you reopen the app, we were grabbing better attention and getting better conversion.”
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
- Sociaaal: Buy apps that went viral and are now declining, then grow them againBought 6-year-old Celebs mid-decline; now bigger than ever, ~$300K+/month
- Sociaaal: Run the same style of test ads for apps you might build or buyStrong ad results count as a key data point in buy decisions; no numbers
- Sociaaal: Rerun old A/B tests that lost once the rest of the app has changedOnboarding paywall removal lost several times, then won big
- Sociaaal: Run about 100 A/B tests a month, starting small and scaling the winnersPortfolio ARPU roughly 3x in 18 months, 2x in 12; about 1 in 10 tests wins
- Sociaaal: Write down the reason behind every A/B test so AI agents can learn from it
- Sociaaal: Make about 4,000 AI video ads a month, with people coming up with the ideasProfitable at 3x LTV/CAC; creative team tripled in 6 months to keep up
- Sociaaal: Keep testing brand-new ad ideas instead of endless versions of one winnerCredited with staying market leader in their niches; no numbers
- Sociaaal: Use AI to find TikToks from other industries to inspire your ads
- WiseSam: Ask ad visitors many questions, reflect answers back, then offer a $1 first weekFounders say the long quiz works very well; no numbers yet
- WiseSam: Give new users one-tap task buttons so the AI assistant learns about them fastUsers get more engaged once the agent knows them; still needs more guidance