Three-tier cascading paywall offers: +8.3% ARPU
Test run bySuperwallTesting and CRO tool
Commentary byLiftArchiveOur analysis, not theirsShowing three tiers of paywall offers by view count (standard, then an extended trial, then a discount) raised ARPU by 8.3% over a two-tier setup, worth about $14.7K a month in proceeds.
Nick Godwin of Superwall ran the test for an unnamed subscription app.
Vahe Baghdasaryan of Tangent shared the results on X.
- ARPU: +8.3% versus the two-tier setup.
- Proceeds: about +$14.7K per month, or about $177K annualized.
- Exact offer terms (trial length, discount size) and conversion rates were not shared.
Control
Cascading offers v1: one paywall for each user's first two paywall views, and a second paywall for users with two or more views.
Variant
Cascading offers v2: the standard paywall on the first view, an extended-trial paywall on views two and three, and a discount paywall from the fourth view on.
Why it likely won
Our reading of the result, not the practitioner's claim.
- Users who skip the first paywall are still interested. A longer trial lowers the risk for them before any price cut.
- Saving the discount for the fourth view protects full price for people who would have paid anyway.
- Each step gives a new reason to say yes, matched to how many times the user has hesitated.
Caveats: The app is anonymized, and the details come from a screenshot in the post. Sample size, duration and offer prices were not shared. Superwall staff ran the test and Superwall sells the paywall tool, so the result may lean positive.
Who ran it
Superwall ran this test. Source: Vahe Baghdasaryan on X (@vahebagdasar). The numbers are theirs; the summary and commentary on this page are ours.
How to run this test on your app
- Hypothesis
- If we show a standard paywall first, an extended trial on the next views and a discount only after several views, then ARPU will rise because each user gets the smallest offer that gets them to convert.
- Primary metric
- ARPU
- Guardrails
- Share of users on discounted price
- Trial-to-paid
- Renewal rate
- Who should test it
- Apps where users hit the paywall many times before paying. If most users see it once, view-count targeting won't have much to work with.
Questions
Should I lead with the discount?
This test suggests not. The winning setup showed full price first and saved the discount for the fourth paywall view.
How long was the extended trial?
The post doesn't say. Test a trial clearly longer than your standard one.
Will repeat discounts train users to wait?
It's a real risk. Watch the share of users who buy at a discount and keep the discount behind several views.
Similar tests
Kovalee