- Home
- Skills
- Marketing & SEO
- Paywall Optimization
More screenshots
Works with the AI Agents you already use
Paywall Optimization
Diagnose a paywall by its weakest funnel stage, fix the screen against three buyer questions, and test one change at a time with honest samples.
$5
Paywall Optimization
Example session with this skill installed
Diagnose our language app's paywall. It appears right after onboarding. Last 30 days: 8,200 installs, 1,900 paywall views, 210 trial starts, 96 still paid after trial. The headline says "Go Premium," plans are monthly at $12.99 and annual at $79.99 with monthly pre-selected, and we show no ratings anywhere.
- Read your context and instructions
- Compiled the paywall optimization
PAYWALL DIAGNOSTIC — language app
Funnel: reach 23% (1,900 / 8,200) → CTA tap 11% (210 / 1,900) → trial→paid 46% (96 / 210). Purchase completion not separately reported; trial starts stand in for it here.
Weakest: engagement scores 0 (11% vs a 25–45% orientation band) and reach also scores 0 (23% vs 60%+ expected for an after-onboarding placement). Per the tie-break rule, engagement is fixed first — users cannot hit a placement problem's remedy until the offer itself converts the viewers already arriving. Trial→paid at 46% scores 2: the trial experience is healthy and is explicitly hands-off.
Screen audit
- Annual exists and undercuts monthly on a yearly basis — the raw material of a good default is present.
- Headline "Go Premium" names a plan, not an outcome. Fix: lead with the result the buyer wants ("Hold real conversations in Spanish") above the plan picker.
- Monthly is pre-selected, anchoring buyers on the worse-value plan and tripling the chances of month-two churn. Fix: pre-select annual, showing the monthly equivalent ("about $6.67/mo, billed $79.99/yr — save 49%") so the low number and the honest total appear together.
- Zero proof above the fold. Fix: add the store rating with review count; a paywall with no proof reads as a demand.
- Trial terms do not state the first-charge date in the material provided. Fix: "Free for 7 days, first charge on [date], cancel anytime," plus the restore-purchase button visible without scrolling.
Ship now: annual as default with the equivalent-price framing; rating line above the fold; trial terms and restore made visible.
Test next: headline only — outcome framing vs "Go Premium" — primary metric CTA tap rate, ~2,000 viewers per variant at the current 11% baseline to detect a 10% relative lift, minimum one full week. Reach is the following cycle's test (after first value moment vs after onboarding), not this one.
Connects securely to your tools. The creator never sees your data.
About this skill
Most paywall work starts with a redesign and ends with nobody knowing whether it helped. This skill starts with the funnel instead. You bring four numbers — reach, CTA tap rate, purchase completion, and trial-to-paid — and a scoring rubric names the weakest stage, because fixing the wrong stage burns a release cycle: polishing plan design when nobody reaches the paywall is placement work, and tuning trial starts while trial-to-paid bleeds is an onboarding problem wearing a paywall costume.
With the weak stage named, the screen gets audited against the three questions a buyer asks in about four seconds: what do I get, what does it cost, and can I back out? Outcome-led headlines, three to five do-phrased benefits, proof above the fold, deliberate default-plan selection with the per-period equivalent and the total both visible, the first-charge date in words, and a visible exit — restore purchase included, because the stores require it and skeptics convert when leaving is easy. A seven-point screen checklist makes the audit repeatable, and a placement table decides when the ask should happen, defaulting to the soft paywall after the first value moment when there is no data to argue otherwise.
Changes ship in two lanes: obvious fixes that need no test, and exactly one experiment per cycle from a leverage-ordered priority list, documented on an experiment card with a hypothesis, the sample size the baseline actually requires, a minimum run, and a verdict rule written before launch. A full worked example (a meditation app whose trial starts were an offer problem, not a trial problem) and a fixed delivery format — funnel line, weakest stage, audit, ship-now, test-next, hands-off — keep every diagnostic consistent.
Edge cases include the no-analytics cold start, a first-ever paywall with no baseline, regulated categories where claim discipline beats conversion advice, and seasonality contaminating a read.
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Marketing & SEO.
- The orientation bands are starting points, not laws; after one measured cycle your own baseline should replace them, and the skill says so.
- Low-traffic apps will find the honest sample sizes slow — the protocol's answer is longer tests, not earlier verdicts.
- It does not implement the paywall, configure the tools, or handle store submission; and trial-retention problems are routed to lifecycle work, not fixed here.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 24 days ago
Permissions
No special permissions declared or detected