Free tool, no login

Review like Pete Koomen

Paste a URL and get your page judged as a conversion experiment: how many steps to the moment your product becomes obviously useful, what friction sits in the way, and whether this page makes a claim you could actually test.

Free, no login, no email. Nothing stored beyond a 24 hour cache. Markdown copy

Review like

Steps to the Aha moment, signup friction, and what you would test first

How it works

  1. 1

    Paste any URL

    We load that one page — its text, its markup, a screenshot, and the type sizes and colours the browser actually resolved. No login and no email.

  2. 2

    Eight tests, and most of them are counts

    Steps to the Aha moment. Required signup fields. Whether the page makes one falsifiable claim. How many audiences the hero serves. Whether the primary action is singular. Whether the value fits above the fold. Whether AI is doing the work or decorating. Whether proof sits where the doubt is.

  3. 3

    Get the one change to ship this week

    The findings come back with the numbers behind them — five fields, four competing buttons, seven steps — and the single change worth shipping first, phrased as the variant rather than as advice to go test something.

Why this matters

Most conversion advice is folklore

Search for how to improve a landing page and you will be told that red buttons outperform green ones, that shorter forms always win, and that social proof belongs above the fold. Some of this was true on somebody else’s page in 2013. None of it is a finding about yours.

What survives is not a list of rules but a habit: count the things that can be counted, name the claim the page is making, and work out what you would change to find out whether the claim is landing. This lens is that habit, applied to one page.

It will not predict a lift, and that is deliberate

The tempting feature here is a number — “this change is worth 14% more signups”. We will not ship it. The tool sees one page. It has no baseline, no traffic, no funnel and no segment data, so any percentage would be produced by a language model writing something plausible, and a fabricated measurement is indistinguishable from a real one once it is in a slide.

So it counts what is actually countable: required fields, discrete steps to value, competing calls to action, distinct audiences addressed in one hero. Those numbers are real, and they are usually higher than the team building the page believes.

The empty dashboard problem

The most common failure this lens catches is a page that has carefully optimised the wrong finish line. Signup is measured, celebrated, and treated as the end of the funnel — and the visitor arrives at an empty dashboard having still not seen the product do anything useful. The Aha moment is further away than the signup button, and pages are rarely designed for the distance between the two.

This is a lens, not a person

Superflow is not affiliated with Y Combinator, and Pete Koomen has not reviewed your page or endorsed this tool. This is our interpretation of published episodes and essays, cited under every result. It does not write as him, and it will not produce a quotation.

Common questions

No, and the tool will not claim otherwise. This is a review lens assembled from the public record: YC's Design Review episodes he appears in, and his own published essays. It never writes as him, never states what he would have said about your page, and never invents a quotation.

The friction you added last sprint is still there

Superflow agents review the pages you ship on every change, against the standards your team actually holds, and leave the findings as comments on the page itself.

Start free