Paste a public URL. Get a Friction Score from 0 to 100 — higher means less friction — and the single biggest structural issue on the page, with the one change to make. No signup, no credit card, no snippet.
That's the whole tool. What follows is what it looks at, and, just as importantly, what it cannot see.
What the checker reads
Value clarity — whether a first-time visitor can tell what's offered and why it matters within the first screen, without scrolling.
Action clarity — whether one action is unmistakably primary. Two buttons of equal weight pointing to different destinations is among the most common and most fixable friction patterns on the web.
Trust and proof — whether reasons to believe sit near the ask rather than in the footer.
Request weight — how much the page asks for relative to what it has given, and whether expensive asks arrive before value.
Information scent — whether each step signals the next.
Mobile viability — whether the page functions as a phone page, not a shrunken desktop one.
First-screen cost — whether load and layout spend attention before anything is read.
Why the score is deterministic
The same page, checked twice, returns the same score. The rules are fixed and hard-coded. No language model sets the number.
That matters more than it sounds. A measurement you can't repeat is a mood, not a baseline. Because this one is repeatable, "change one thing and re-check" is a valid way to learn something: if the score moved, your page moved it, not the measurement.
What to do with the score you get
Treat the first number as a baseline, not a verdict. Its usefulness begins on the second run.
Make the one change the checker names. Re-check. Because the measure is deterministic, a movement in the number is a movement in the page — not a different opinion on a different day. That's the entire value of repeatability, and it's why we don't let a language model choose the number.
If the score doesn't move, that's informative too: the issue you fixed was real but lighter than the next one down the ranking.
What the checker cannot tell you
Worth saying plainly, because most tools in this category don't.
It reads structure, not behaviour. It can tell you a form is heavy. It cannot tell you which field your visitors stop at.
It has no visitors. It doesn't know which step of your funnel loses the most people, because that lives in what real people do on your real pages.
It is a friction measure, not a conversion prediction. Friction is one of several things determining whether people convert. Any tool promising "raise this number by 10 for X% more revenue" is asserting something it cannot know.
A score tells you what. A flow tells you where. The checker does the first honestly, and the tracking snippet — installed later, if you want it — does the second.
Common questions
One score and one fix per URL, with no account. The paid product is the continuous, behavioural half.
There isn't a universal pass mark, and any tool quoting one is guessing. Use your first score as a baseline and move it.
Not for the check. It reads a public URL.
No. Anything requiring authentication is invisible to an external scan — which is precisely the boundary the snippet exists to cross.