News-style paywall vs classic paywall: +190% direct conversion
Commentary byLiftArchiveOur analysis, not theirsTipStop saw a 190% jump in direct conversion when it tested an "NYT"-style paywall against its "Classic" one.
TipStop, a sports betting tips app, ran the test with Purchasely's A/B testing tool, and the result was published on the Purchasely blog. The post does not describe either design.
- Direct conversion: +190% for the NYT-style paywall over the classic one.
- Sample size, duration and confidence were not shared.
- Revenue, trial-to-paid and retention were not reported.
Control
TipStop's "Classic" paywall. Its layout was not described.
Variant
A paywall in the style of the New York Times. The post does not say what that meant in practice.
Why it likely won
Our reading of the result, not the practitioner's claim.
- We can't say why it won, since neither design is described. A news-style paywall often shows some content first and then asks to subscribe, but that is a guess here.
- A very large lift like this can also come from a small sample or a weak baseline, so treat it as a lead, not a benchmark.
Caveats: Retold in a short vendor post by Purchasely, which sells the testing tool used, so it may lean positive. No designs, sample size or duration were shared.
Who ran it
TipStop ran this test. Source: Purchasely blog. The numbers are theirs; the summary and commentary on this page are ours.
How to run this test on your app
- Hypothesis
- If we switch from a classic plan-list paywall to a news-style paywall, then direct conversion will rise, because the layout fits how users of a content app expect to subscribe.
- Primary metric
- Direct conversion (paywall view to purchase)
- Guardrails
- Refund rate
- Renewal rate
- Revenue per user
- Who should test it
- Content and tips apps where users come back daily for fresh material. Utility apps with no ongoing content may not benefit from a news-style approach.
Questions
What is an NYT-style paywall?
The source does not define it. The term usually refers to the paywalls news sites use, but check the actual design before copying it.
Is +190% a realistic target for a paywall redesign?
Treat it as an outlier. With no sample size or baseline shared, it is hard to judge how much is real lift and how much is noise.
Similar tests
Kovalee