Paywall Psychology: The Tier Structures That Actually Convert Free Viewers
Evidence from 200+ checkout experiments — why three tiers beat five, how anchoring moves average revenue per user, and the trial mechanic that doubles day-30 retention.
Ask a room of product managers how many subscription tiers a media product should offer and you will get five different answers. Ask the data, and you get one: three.
Across the checkout experiments we’ve catalogued — over 200 controlled tier-structure tests spanning publishers, video platforms, and creator networks — three-tier architectures consistently outperform both two-tier (binary) and five-tier (menu) designs by 20–35% on revenue per visitor.
Why Three Works: The Decoy in the Middle
The mechanism is classic asymmetric dominance. A three-tier layout lets the platform position the middle tier as the obvious choice — framed on one side by a deliberately austere entry tier and on the other by a premium anchor whose price exists primarily to make the middle look reasonable.
| Structure | Median Conv. Rate | ARPU Index | Notes |
|---|---|---|---|
| Single price | 1.9% | 100 | No choice architecture |
| Two tiers | 2.4% | 118 | Binary framing splits audience |
| Three tiers | 3.1% | 147 | Anchored middle wins |
| Five tiers | 2.2% | 109 | Choice paralysis sets in |
The Mechanics That Compound
- Price anchoring reads right-to-left in impact: a $24 premium tier lifts middle-tier conversion by ~30% even if it only sells to 3% of members.
- Annual discount framing beats percentage framing: “2 months free” outperforms “save 17%” at checkout by a consistent margin — concreteness beats arithmetic.
- The trial that retains: 7-day trials with a calendar reminder (“your trial ends Tuesday”) retain 2.1x better at day-90 than frictionless trials that auto-convert silently. Reminders feel like they should leak cancellations — they actually build trust equity that survives the first billing cycle.
“Nobody wants to buy the cheap option or the expensive option. Design the middle as the identity: this is what a member like you chooses.”
The complete experiment library — including checkout friction tests and annual/monthly toggle placement — lives in The Paywall Conversion Playbook.
Tiers are not a menu. They are an argument.