A subscription app · B2C · App Masters client
Show 2-3 screens of benefits and testimonials right before the paywall
Won the client's test against the other onboarding version
Workedonboardingpaywall
What they did
Host's recommended flow, tested for an App Masters client: before an unchanged paywall, add two or three screens: an inviting benefit headline (e.g. 'Level up your painting game'), the features users already say they love, and social proof/testimonials, with a Continue button, then the paywall. The alternative version tested did not work; this one won.
What happened
This version won the A/B test; no numbers
- Stage
- activation
“All right, just adding these two, three screens before your paywall. Keep the paywall the same.”
“Put some social proof. You got amazing social proof. You got amazing testimonials. I know you have it in your paywall, but lead them up to it. And then, hit them with the paywall.”
“It did not work for a client of ours. This boop boop did not work. This one won.”
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
- Value Study: Name your app after the term people already learn in classSteady users for years with no marketing; trade-off is a hard-to-brand name
- Value Study: Rank for many small, specific keywords that only your real users search1,000+ keywords ranked; low volume but high conversion
- Value Study: Show what the app does in the first screenshot, not a menu
- Value Study: Win over teachers, who hand your app to a whole class at onceMain growth driver; teachers share with ~30 students at a time; founder says paid outreach to them struggled
- Value Study: Sell school licenses through a separate schools-only appDidn't work; no school took it
- Value Study: Turn the workarounds users show you into built-in featuresFounder says it worked; no numbers
- Value Study: Remove a feature people call confusing once a newer one covers itBackfired; many users who relied on it complained
- Value Study: Ask an optional 'Where did you hear about us?' with a free-text box31,000 answers; free text names specific teachers, courses and books
- Value Study: Ask which mediums users work in so the app feels made for them
- Value Study: Translate your app into several more languagesPaying users in the new languages; Dutch became common out of nowhere
- Value Study: Add a one-time Pro unlock to an app that has always been freeAbout £200 a month, steady; most users stayed free and missed key features
- Value Study: Announce 3 months ahead that the free tier ends; lifetime buyers keep accessClear spike in lifetime purchases before the switch; no review backlash mentioned
- Value Study: Drop the free tier and charge everyone a small amountSubscriptions and lifetime sales rose; revenue doubled in a year to $61K
- Value Study: Offer a lifetime plan next to the yearly plan50% lifetime, 50% yearly until the Android launch
- Value Study: On Android, replace the hard paywall with a soft paywall trialAndroid reviews recovered from very bad
- Value Study: Reply personally to every support email and fix bugs right awayMany reviews name the founder; turned angry users into five-star reviewers
- Value Study: Hire a cheap influencer with a big following in another language$4,000 in one week and a lasting lift; broken Android translations brought bad reviews
- Value Study: Run an ad in the main print magazine for your hobby nicheRoughly broke even; reached people who wouldn't have found it
- Value Study: Run Meta ads for a niche hobby appNo positive ROI yet; only dabbled
- Value Study: Add a tip jar to your settings menu$157 in tips this year, covering the developer fee
- Value Study: Test higher prices in the US firstSettled on a price about halfway up; no numbers
- Rank number one for your niche's main keywordsAbout $30K MRR; #1 for ~10 related terms
- Add Meta ads once organic growth in your niche levels offRevenue doubled or tripled; took time
- Raise your yearly price from $29.99 to $49.99Good so far, per a viewer; no numbers