Practitioner pattern · Vahe Baghdasaryan (Sells paywall optimization services (Tangent); all paywalls shown are built on Superwall)
Add paywalls at app open, after scans, and on cancel before testing designs
Guest's first step for every new client; no numbers
Workedpaywallexperimentation
What they did
Audit and add the missing placements first, then start A/B testing: transaction abandonment; session start (at least once per day); after every X core actions (e.g. every N scans in a coin identifier); freemium usage limits ('you are almost out of your 5 free scans, subscribe for unlimited'); an offer after a trial cancel; another after a subscription cancel. Placements are the foundation, not the experiment.
What happened
Guest says most new clients lack these and adding them is the baseline for gains.
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
- Claim: Say 'free' 5-7 times across a 3-screen paywall before showing the price34% trial start rate; multi-step paywalls lift conversion 20-40% vs one screen
- Call the trial plan 'Try it free' instead of 'Annual', showing the price belowUp to ~10% more abandoned purchases, but a net gain if trial-to-paid is above 30%
- Claim: Drop the free trial once trial starts and trial-to-paid are both strongMore revenue without the trial, guest says; no numbers given
- Cal AI: Show a spin wheel that lands on 80% off instead of just showing the discountDouble-digit conversion lift vs a plain discount screen in an A/B test
- Rizz: Let users skip the free trial for a lower price: $6.99 vs $9.99 a week
- Cal AI: Let users turn the free trial off, at the same price, for high-intent appsSome users buy right away, guest says; no numbers
- Quitter: Show no prices on the 3-page paywall until users tap 'Start my journey today'The biggest win on this paywall, guest says; no numbers
- When a purchase is cancelled, offer a discount on the annual plan onlyWeekly or monthly discounts failed in dozens of tests
- On the abandoned-purchase offer, give 80% off with a trial or 90% off without~10% of buyers skip the trial and pay right away
- Fix paywalls for five months while installs stay flat, with no ad spendInstall-to-paid rose from 1.1% to ~4% in five months
- Tangent: Borrow discount and urgency tricks from Temu and Shein for paywall testsSource of the spin wheel and 'limited spots' ideas; guest says it works