Launching today
Zaplens
Simulate real users and fix your UX 100x faster & cheaper
16 followers
Simulate real users and fix your UX 100x faster & cheaper
16 followers
Conventional A/B and CRO tools require heavy traffic to reach statistical significance, so most sites never conclude a test. Zaplens removes that dependency: 100+ synthetic AI personas traverse a URL toward a goal and return results 100x faster and cheaper. It surfaces UX bugs without treating your visitors as lab rats, with reasoning, so you know exactly what to fix. Before *any* your users see it. Choose personas from Zapstore or bring your own, but make sure you Zap it, before you launch it!
👋 Hey Product Hunt! I'm Abhishek, the maker of Zaplens.
Here's a feeling every product team knows: you've got two versions of a checkout, or a redesign sitting in a PR, and you genuinely don't know which one to ship. UX isn't like the rest of your code, you can unit-test a function, but there's no test for "is this page actually usable?" So you ship it and hope.
The usual fix is an A/B test, but that's four to six weeks of waiting for significance that often never comes. For a team that loves to ship fast, that's a parking brake. And the newer AI tools just hand you a verdict from a black box, "this page is confusing." Okay… why? Says who? I could never trust a recommendation I couldn't see the reasoning behind.
So I built Zaplens to fix all three. Drop in a URL, set a goal, and 100 to 1,000 AI personas walk through your site. About 90 seconds later you get a heatmap of where they got confused, what drew them in, and where they bailed — and every dot is pinned to that persona's exact reasoning, so you always see the why. Run two versions head-to-head, bring your own personas, run it from your favorite coding tool (Claude Code, Codex etc) you already work in. It's the test suite UX never had and its faster and far cheaper than a six-week study.
Somewhere in the beta, "zap it before you ship" became a phrase people started saying on their own. That's when we knew it was ready, and we're so excited to launch it today.
Point it at your most-suspect page and tell us what surprised you.
→ zaplens.site
— Abhishek
mailX by mailwarm
How do you validate that the persona behavior matches real users, not just what the model thinks users do?
@naimz Great question! we have evaluated grounded personas which matched with real world users to see how closely we could mimic their behaviour. We plan to scale this evaluation framework to many different personas in the coming weeks.