Media · Growth
Metered paywall A/B test
Testing how many free articles to give before the paywall, with return visits as a guardrail.
Paywall: 3 free articles vs. 5
+21.7%
Significant winSubscription starts per reader
095% interval +9.4% to +34%
3 free articles
5 free articles
control
Who it's for
Subscription and growth teams at publishers who test paywall rules and need to know what they cost in audience.
What it solves
Analyze the paywall test on the exposure data in your warehouse, with the same subscription definition as your reporting, and see whether the extra subscribers cost you readers.
Questions it answers
Did 3 free articles lift subscription starts?
Subscription starts rose from 1.20% to 1.46% of readers, a 21.7% lift (95% interval +9.4% to +34.0%).
Did return visits or ad impressions fall?
Return visits per reader fell 1.2% and ad impressions 0.8%, both inside their 3% guardrails.
- Is the lift significant, and how wide is the interval?
- Which traffic sources responded best?
The impact
- Set the paywall meter on evidence, not on instinct
- Weigh new subscribers against lost ad audience
- One subscription definition for tests and reporting
The data it needs
- Exposure events from your experimentation or paywall tool
- Article view and paywall events
- Subscription start events
Read in place in your warehouse. Nothing is copied out.
Metered paywall A/B test: frequently asked questions
Run this on your own data.
Start free on your warehouse, or book a demo and we'll show it on your data.