paywall-upgrade-cro
by coreyhaines31paywall-upgrade-cro helps teams improve in-product paywalls, upgrade screens, trial-expiration prompts, and feature gates. Learn when to use it, how to install it, which files to read first, and how to apply it to freemium, trial-to-paid, and tier-upgrade conversion work.
This skill scores 78/100, which makes it a solid directory listing candidate: agents get a clearly scoped trigger surface for in-product paywalls and upgrade prompts, plus enough structured guidance to produce better recommendations than a generic CRO prompt. Directory users should still expect a document-driven skill rather than a fully operational workflow, with limited implementation assets beyond written frameworks and experiment ideas.
- Strong triggerability: the description explicitly covers paywalls, upgrade screens, upsells, feature gates, trial expiration, and free-to-paid conversion moments, and distinguishes this from pricing pages and pricing strategy.
- Good operational guidance: SKILL.md includes an initial assessment, asks the agent to check for product marketing context files first, and the evals show concrete expected behaviors like handling usage limits, emotional state, paywall components, and anti-patterns.
- Useful leverage for optimization work: the included experiments reference provides a substantial A/B testing menu across timing, design, pricing presentation, copy, personalization, and UX frequency decisions.
- Adoption is mostly content-only: there are no scripts, install steps, rules, or reusable templates, so execution depends on the agent interpreting a long markdown guide correctly.
- Practical outputs are implied more than packaged: the evidence shows frameworks and expected recommendations, but not a concise quick-start deliverable format or explicit step-by-step workflow artifacts for handoff.
Overview of paywall-upgrade-cro skill
What paywall-upgrade-cro does
The paywall-upgrade-cro skill helps you design and improve in-product upgrade moments: paywalls, upgrade screens, upsell modals, trial-expiration prompts, and feature gates. It is built for conversion work inside the product, not for public pricing pages or broad pricing strategy.
Who should use this skill
Best fit readers are founders, PMs, growth teams, product marketers, and designers working on freemium-to-paid, trial-to-paid, or tier-upgrade flows. If users already reached some value and now need a reason to upgrade, paywall-upgrade-cro is the right tool.
The real job-to-be-done
Most teams do not need abstract CRO theory. They need a usable upgrade screen that matches the trigger moment, the user's emotional state, and the product's actual value. This skill is strongest when you want concrete screen structure, message direction, and test ideas rather than generic "improve conversions" advice.
What makes it different from a generic prompt
The skill is opinionated about context first. It pushes the model to identify the upgrade context, understand the product model and user journey, and tailor the paywall to the trigger type. The repository evals show it expects details like value framing, plan comparison, social proof, CTA quality, and escape hatches instead of shallow copy rewrites.
What it is not for
Do not use paywall-upgrade-cro for Conversion when the real problem is pricing architecture, package design, or a public landing-page pricing table. The skill itself points to separate approaches for pricing strategy and page CRO. It also will not replace product analytics; it improves the upgrade moment, but it cannot invent missing funnel data.
How to Use paywall-upgrade-cro skill
Install paywall-upgrade-cro in your skills setup
Use:
npx skills add https://github.com/coreyhaines31/marketingskills --skill paywall-upgrade-cro
If your environment already syncs the full repo, confirm the skill exists at skills/paywall-upgrade-cro/ and that SKILL.md is available to the agent.
Read these files first
Start with:
skills/paywall-upgrade-cro/SKILL.mdskills/paywall-upgrade-cro/evals/evals.jsonskills/paywall-upgrade-cro/references/experiments.md
This reading order matters. SKILL.md gives the workflow, evals/evals.json shows what good outputs must contain, and references/experiments.md is useful once you need test backlog ideas rather than first-pass diagnosis.
Check for product marketing context before prompting
A key implementation detail in this repository: the skill explicitly tells the agent to check for .agents/product-marketing-context.md or .claude/product-marketing-context.md before asking questions. If you already maintain one of those files, this paywall-upgrade-cro skill becomes more useful because it can ground recommendations in your positioning, ICP, and value props.
Know the input the skill actually needs
For strong output, provide:
- upgrade type: freemium, trial expiry, tier upgrade, feature upsell, usage limit
- trigger moment: what action causes the paywall
- current free vs paid boundary
- target user segment
- current screen or copy
- current conversion rate, if known
- user friction or objections you already hear
- device/context: web app, mobile app, modal, full screen, settings page
Without these, paywall-upgrade-cro usage tends to stay generic.
Turn a vague request into a complete prompt
Weak:
“Help me improve our paywall.”
Strong:
“We run a project management SaaS. Free users can create 3 projects; the 4th project attempt triggers a modal. Paid unlocks unlimited projects, custom fields, and team permissions. Current upgrade rate from this modal is 1.8%. Users are usually solo operators at this point and feel blocked. Rewrite the upgrade screen using the paywall-upgrade-cro framework. Include headline, supporting copy, plan comparison, objection handling, CTA, and one escape hatch. Also suggest 5 experiments.”
The stronger version gives the skill enough to adapt the message to the moment instead of writing generic upsell copy.
Match the paywall to the trigger type
The repository evidence shows the skill distinguishes trigger contexts, and that changes output quality:
- usage limit: user is frustrated and trying to continue
- trial expiration: user may need reminder of achieved value
- feature lock: user needs relevance and justification
- tier upgrade: user needs plan-fit clarity, not beginner education
If you do not specify the trigger, the model may design the wrong kind of screen.
Ask for the full screen structure, not just copy
The evals indicate better outputs include core components such as:
- headline tied to value, not just restriction
- proof or demonstration of what paid unlocks
- free vs paid comparison
- social proof or trust signal
- specific CTA
- escape hatch or dismissal path
That is a better way to use paywall-upgrade-cro guide workflows than asking only for “better text.”
Use the experiments reference the right way
references/experiments.md is broad and practical. It is most useful after you have a baseline recommendation and want testable variants across:
- trigger timing
- hard gate vs soft gate
- layout
- value presentation
- pricing display
- CTA and copy
- personalization
- frequency and dismissal behavior
Do not start with a huge experiment list if your current paywall is fundamentally unclear. Fix message-market fit first, then expand the test plan.
Repository signals that help you trust the skill
The eval file is especially valuable for install decisions because it reveals what the agent is supposed to do in realistic paywall scenarios. It expects the skill to identify emotional state, screen components, anti-patterns, and concrete copy recommendations. That makes paywall-upgrade-cro install more attractive than skills that only contain abstract principles.
Practical workflow for real teams
A good operating sequence is:
- Share your current trigger and free/paid boundary.
- Give the current screen copy or screenshot text.
- Ask the skill for a revised structure and copy.
- Ask for anti-patterns in your current implementation.
- Ask for 3 to 5 experiments from
references/experiments.md. - Turn the best output into a product/design spec.
This workflow gets more value from paywall-upgrade-cro for Conversion than treating it like a one-shot copywriter.
Common adoption blockers
Teams usually stall for one of four reasons:
- they have no clear free-to-paid boundary
- they ask for a paywall before defining the trigger moment
- they want pricing strategy, not screen optimization
- they provide no product context, so outputs feel interchangeable
If that sounds like your situation, fix the inputs before judging the skill.
paywall-upgrade-cro skill FAQ
Is paywall-upgrade-cro good for beginners
Yes, if you can describe your product and upgrade moment clearly. The skill is structured enough to guide non-experts toward a usable paywall, but beginners still need to provide the basics of what is free, what is paid, and when the ask appears.
Is this better than a normal prompt
Usually yes for this specific use case. A normal prompt may produce surface-level copy. The paywall-upgrade-cro skill is stronger because it is scoped to in-app upgrade moments and pushes for context, user state, screen components, and test ideas.
Does it help with pricing strategy
No, not directly. It helps present and optimize an upgrade ask inside the product. If your core problem is price point, packaging, or monetization model, this is the wrong first skill.
Can I use it for web and mobile apps
Yes. The repository scope covers in-product upgrade prompts broadly. Just specify the UI format because a mobile paywall, modal, and desktop upgrade page need different density and hierarchy.
When should I not use paywall-upgrade-cro
Skip it when:
- you need a homepage or pricing page rewrite
- users have not yet experienced product value
- you are solving retention or onboarding, not upgrading
- you have no meaningful product context to supply
Does it include experiment ideas
Yes. The references/experiments.md file is one of the strongest support assets in this skill and is useful for building a real testing backlog after initial recommendations.
How to Improve paywall-upgrade-cro skill
Feed it better evidence, not bigger prompts
The biggest improvement lever is better context quality. Include:
- actual trigger event
- user segment at that point
- what they just tried to do
- what paid unlocks right now
- conversion baseline
- top objections from sales, support, or research
This improves output far more than asking for “10 variants.”
Provide the user's emotional state
One subtle strength in the evals is attention to user emotion at the trigger. If a user just hit a hard limit, they are blocked and impatient. If a trial expired, they may be uncertain whether the product earned payment. Tell the skill how the user likely feels; the copy and CTA quality usually improves.
Ask for anti-pattern detection explicitly
A high-value prompt addition is:
“List the anti-patterns in our current paywall before rewriting it.”
That often surfaces issues like:
- leading with restriction instead of value
- vague CTA text
- too many plan choices
- no justification for timing
- no escape hatch
- no proof the upgrade is worth it
Use screenshots or exact copy when possible
If you can paste your current screen text or describe the exact layout, paywall-upgrade-cro usage becomes much more actionable. The skill is more effective at transformation than blind invention.
Iterate in two passes
Best practice:
- Pass 1: diagnose the trigger, message, and friction
- Pass 2: generate revised copy, structure, and experiments
This prevents the common failure mode where the first answer jumps to polished copy without first checking if the screen is solving the right problem.
Pull experiment ideas after the baseline is solid
The experiments reference is useful, but easy to misuse. Improve the skill's output by first asking for the best single paywall for your current situation, then asking for variants to test. Otherwise you get a long list of experiments without a strong control.
Compare outputs against the eval expectations
Before shipping, sanity-check the answer against what evals/evals.json implies strong outputs should include:
- correct trigger classification
- value-led headline
- clear paid benefits
- comparison or choice clarity
- strong CTA
- emotional-state awareness
- escape hatch
- anti-pattern warnings
That is one of the easiest ways to get more reliable results from paywall-upgrade-cro.
Add your own internal context file
If you use this skill repeatedly, create .agents/product-marketing-context.md with your ICP, pains, positioning, proof points, and plan structure. This reduces repetitive prompting and makes the skill more consistent across projects.
Know the limits of paywall-upgrade-cro for Conversion
If the output still feels weak after good inputs, the issue may not be the skill. Common root causes are weak product value, poor activation, unclear packaging, or low-intent traffic. paywall-upgrade-cro can improve the upgrade moment, but it cannot fix upstream product-market or pricing problems on its own.