Skip to main content

Experiment on your paywall

Find the best paywall scientifically 💙.

Start now with this prompt into your Claude Code or Codex

I want to use Experiwall (experiwall.com) to A/B test my app. First, analyze my project. Look at the codebase, figure out what kind of app it is (B2B, B2C, SaaS, e-commerce, etc.), and recommend which pages or UI elements I should be experimenting on: pricing pages, onboarding flows, CTAs, upgrade prompts, sign-up forms, feature gates, whatever makes sense for my product. Then set up Experiwall in my codebase. Prefer the Experiwall CLI over the MCP server: start with npx @experiwall/cli login. If the CLI needs browser authentication or an API key, open https://experiwall.com/cli/login directly for me so I can sign in and generate the CLI login command. After login, run npx @experiwall/cli whoami or the equivalent command to verify authentication. Use the CLI for setup, experiment creation, and verification whenever possible. Only connect to the Experiwall MCP server if the CLI is unavailable or you need AI-native project management. The MCP endpoint is https://experiwall.com/api/mcp. For most websites, install the lightweight Experiwall script snippet and add conversion goals or events where needed. Only use @experiwall/react if this app needs React component-level A/B tests. Create my first A/B experiment with at least two variants based on your analysis. I want real conversion tracking so we can see which variant actually performs better. When you need me to open a website or preview the app, use the appropriate command to open the browser directly so it is easy for me to review. Set everything up end-to-end so it's ready to go live.

Compatible with

Claude
Codex
Cursor
Windsurf
VS Code
Kiro
Lovable
OpenCode
Gemini CLI

Find the best design with experiments

Describe what you want to test, and Experiwall sets up multiple variants ready to A/B test.

Your prompt

"Create a premium upgrade page for a fitness app. Highlight the yearly plan with a 40% discount badge. Use social proof and a free trial CTA."

Generating...
Variant A
Yearly-40%

$59.99/year

Monthly

$9.99/month

50,000+ users
Variant B
Best ValueSAVE 40%

$4.99/mo

billed yearly at $59.99

Monthly

$9.99/month

Start 7-day free trial

Lab Results

=Users who converted
=Users who engaged
=Users who dropped
variant_a(control)

50% traffic

User dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser converted — purchased or subscribedUser converted — purchased or subscribedUser converted — purchased or subscribedUser converted — purchased or subscribedUser converted — purchased or subscribedUser converted — purchased or subscribedUser converted — purchased or subscribedUser converted — purchased or subscribedUser converted — purchased or subscribedUser converted — purchased or subscribedUser converted — purchased or subscribedUser converted — purchased or subscribed

12.4%

conversion rate

Users exposed4.8K
Users scrolled60.0%
Users clicked CTA1.4K
Users bounced25.0%
Revenue$10490.16

baseline

variant_bWinner

50% traffic

User dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser dropped — left without interactingUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser engaged — scrolled through the screenUser converted — purchased or subscribedUser converted — purchased or subscribedUser converted — purchased or subscribedUser converted — purchased or subscribedUser converted — purchased or subscribedUser converted — purchased or subscribedUser converted — purchased or subscribedUser converted — purchased or subscribedUser converted — purchased or subscribedUser converted — purchased or subscribedUser converted — purchased or subscribedUser converted — purchased or subscribedUser converted — purchased or subscribedUser converted — purchased or subscribedUser converted — purchased or subscribedUser converted — purchased or subscribedUser converted — purchased or subscribedUser converted — purchased or subscribed

18.7%

conversion rate

Users exposed4.8K
Users scrolled68.0%
Users clicked CTA2.0K
Users bounced18.0%
Revenue$16327.08

95% confidence

We track every purchase so you know where to invest

From a single hook to production experiments. Experiwall gives you the tools to test, measure, and ship with confidence.

Automatic Bucketing

Deterministic, seed-based assignment. Users always see the same variant. Works client-side and server-side.

Conversion Tracking

Every purchase, signup, and click is tracked per variant automatically. You always know which version makes more money.

Why I built this

Hello! I'm Marcelo

Website problems don't wait for you to find them. A broken link loses you a sale. A missing meta tag drops your ranking. Your website not loading fast enough costs you customers. And you don't notice until it's already cost you. 🥲

I know you can fix it, but it takes a lot of time to keep checking the website every day. Identifying problems and opportunities.

I kept running into this: something would break or underperform and I'd only find out weeks later, by accident. I had no system telling me what to fix and what to improve, every single day.

So I made Experiwall. It runs automatic daily checks on your website for SEO gaps, accessibility issues, performance problems, and how AI tools like ChatGPT talk about you vs. your competitors. Every finding is ready to highlight, copy, and paste straight into your AI agent to fix or improve it immediately. 🥳

We know you use AI to build your product, so we send you prompts that you can copy and paste to fix issues and improve it daily. Of course, under your control. That means, tomorrow you'll have a better product, with less time and tokens spent trying to fix it yourself. ✅

And when you're not sure which version of a page works better, Experiwall runs A/B experiments for you. We help you show two versions of a page to real users, see which one gets more signups or purchases, and choose the best one.

We save you time and tokens

One email each morning. Your product improving every day. Ready to make it better daily? 😊💙

Start Free Trial

Pricing

Simple, transparent pricing

Start free. Upgrade when you need more.

Free

$0/mo

 

Get started free
  • 1 website monitored
  • 50k events / month
  • 1 active experiment
  • 30-day data retention
  • SEO & performance monitoring
  • Web SDK
Most Popular

Pro

$19/mo

Billed $190/yr

Everything in Free, plus:

  • 3 websites monitored
  • 500k events / month
  • 10 active experiments
  • 90-day data retention
  • AI-ready fix prompts
  • 1 team member
  • Webhooks & API access
  • MCP server

Growth

$39/mo

Billed $468/yr

Everything in Pro, plus:

  • 10 websites monitored
  • 5M events / month
  • Unlimited experiments
  • 180-day data retention
  • Experiment templates
  • 5 team members
  • Priority support

Scale

$119/mo

Billed $1,428/yr

Everything in Growth, plus:

  • Unlimited websites
  • 50M events / month
  • 1-year data retention
  • Custom integrations
  • Unlimited team members
  • 99.9% uptime SLA
  • Dedicated support

FAQ

Frequently asked questions

What is Experiwall?

Experiwall is an A/B experimentation platform for websites and paywalls. Add one script tag and you can run split tests, track real conversions, and find which variant generates more revenue.

Which platforms does Experiwall support?

Any website. The experimentation script works with HTML, WordPress, Webflow, Squarespace, Astro, Next.js, and any platform that supports a script tag. No npm install or framework dependency required.

How long does it take to set up an experiment?

Under 5 minutes. Copy one script tag, paste it before the closing body tag, and your first experiment can be live immediately. We also give you a ready-to-go AI prompt that handles the full setup in any framework.

How does variant bucketing work?

Visitors are assigned to variants deterministically using a seed-based hash. The same visitor always sees the same variant across sessions — no cookies required. Works client-side and server-side.

How does conversion tracking work?

Every purchase, signup, and click is tracked per variant automatically. You see which variant generates more revenue, not just more clicks. No manual event wiring required.

Which payment providers are supported?

Hosted paywalls support Stripe and Polar.sh. With Stripe, you paste your own API keys and funds flow directly to your Stripe account — Experiwall never touches the money. With Polar.sh, you connect via OAuth in one click; Polar acts as the Merchant of Record so you skip the Stripe setup entirely. You pick one per project and can switch at any time.

Does the tracking script use cookies?

No. Experiwall is fully cookieless. Visitors are identified server-side using an IP + User-Agent hash. No consent banner required, no cookie storage, and no personal data leaves your visitors' browsers.

What is 'Copy to LLM'?

Copy to LLM gives you a ready-made prompt that includes full context about your experiment — variants, metrics, and targeting rules — so you can paste it into Claude, ChatGPT, or any AI tool and get instant help optimizing your test.

How much does Experiwall cost?

One plan, one price. Creator is $9/month (or $99/year, 1 month free) with a 7-day free trial. You get 10 experiments, real-time analytics, and advanced targeting. No per-seat charges, no usage-based surprise billing.

Can I exclude myself from experiment tracking?

Yes. Open the browser console on your site and run localStorage.setItem('experiwall_ignore', 'true'). The script will stop tracking your visits. Localhost is also ignored by default.

What kind of support do you offer?

You get direct support from Marcelo, the founder who builds the product. No chatbots, no ticket queues. Reach out anytime and expect a response within an hour.

Ready to grow your app revenue?

Start running experiments in your code today. One hook, zero config. 7-day free trial.

Start Free TrialTalk to Marcelo