Subscription Strategy & Paywall Design

This skill exists to stop: locking the one feature that solves the user's pain behind the paywall (so free users never see value), and selling a once-a-month use case as a monthly subscription.

Quick start

Paste the current plan/pricing table and ask: "Which model and which revenue-tree lever?" You get: freemium / trial / reverse-trial recommendation with reasoning, the hero-feature check, a three-tier table mock, the exact tree branch + measuring event, and one single-variable A/B.

🤖 0. HOW TO USE (agent workflow)

A. Choose the model: freemium / free trial / reverse trial. Default recommendation for products whose value needs to be felt before it is understood: reverse trial (full access, then step down). B. Audit a pricing page: rule of three, center-stage target tier, anchor tier, one highlighted "recommended" — then check honesty (no fake urgency, 1-click cancel). C. Locate the lever on the revenue tree (§3) before touching anything: which branch does this change move, by how much, measured by which event? D. Frequency check: if the core use case is A30 (monthly), add A1/A7 companions before scaling paid acquisition. Standard output: model choice + why · pricing table mock · revenue-tree node targeted · one A/B with a single variable · churn hypothesis.

1. Reverse trial vs freemium

Freemium's two blind spots: (1) locking the hero feature — the team guesses which features are "advanced" and locks the one that actually kills the pain, so free users conclude the product is useless; (2) free forever — free is so generous nobody upgrades. Reverse trial: sign up with email only → 100% access for N days (14–15 is the common window) → user invests time and data, reaches the aha moment → auto-downgrade → loss aversion does the selling. The user stays in the ecosystem either way.

The signup gate is half the mechanism, and it is easy to miss: a classic free trial asks for a card before the user has felt anything, so the cost of trying is a payment decision. A reverse trial asks for an email. Same trial length, different population walking in — compare drop-off at the signup step before attributing a conversion difference to the paywall copy.