What is your website doing to the person reading it?
Most website checks tell you what a crawler thinks. This one asks a different question: how much work is your page making a stranger do before they decide whether to trust you. Fourteen checks, each one explaining the behaviour behind it, and no email to hand over.
Reads the public HTML of one page, the way any visitor's browser does. Five checks an hour.
What the fourteen checks are
Every one of them reports the number it found, not a colour. If you disagree with a verdict you can still see what it counted.
- One destination, several namesCounts how many different wordings link to the same page. Four names for one button split a visitor's recognition four ways.
- Trust signalsHTTPS, a privacy policy, a contact address, a phone number, social proof. A stranger looks for cheap proxies before they look at your copy.
- The first sentenceWhether there is one main heading and whether it says what you do in twelve words or fewer.
- Sentence lengthAverage words per sentence. Comprehension falls away past about twenty, and faster on a phone.
- Image descriptionsImages with no alt text. Both an accessibility question and the only way a search engine knows what a picture shows.
- Layout stabilityImages with no width and height. Without them the page jumps as each one arrives, which Google measures as Cumulative Layout Shift.
- Built for a phoneWhether the viewport is declared. Without it a phone shrinks the desktop layout and the text arrives unreadable.
- Your search resultTitle and description length against what Google actually renders, roughly sixty and a hundred and sixty characters.
- Weight before first paintHTML size, scripts and stylesheets. A visitor reads a slow page as a careless business before reading a word.
- A way to replyWhether there is a form, and how many fields it asks for. Every field is one more chance to stop.
- Heading ladderHeadings that jump a level. A screen reader moves by heading the way a sighted reader moves by skimming.
- Link textLinks that say only "read more" or "click here". A link is a promise about what is on the other side.
- Social previewog:image, og:title, twitter:card. Whether a pasted link arrives as a picture or as a bare blue URL.
- Answer-engine readinessStructured data blocks and their types. Search is turning into answering, and answers are read from JSON-LD, not paragraphs.
Questions
What the check reads, what it keeps, and what it cannot see.
Is this an SEO checker?
Not really. An SEO checker asks what a crawler thinks of your page. Most of these fourteen checks ask what a person does when they land on it: how many decisions you put in front of them, how hard the writing is, whether a stranger is given any reason to believe you are real. Four of the checks overlap with SEO because the same things happen to matter to both.
What does it actually read?
The public HTML of the one address you paste, once, the same way any visitor's browser fetches it. It does not log in, does not follow your internal links, and does not run your JavaScript. If your page builds itself entirely in the browser, this will say so rather than guess.
Do you keep my URL or my results?
No. Nothing is written to a database and there is no email field. The report exists for as long as that page is open in your browser.
Why did a nearly empty page score well?
It should not, and an earlier version of this did exactly that. A check with nothing to measure now returns "not applicable" and is left out of the score instead of quietly counting as a pass. A page with no images cannot earn marks for image quality.
The score seems harsh. Is it wrong?
The score is the blunt part. The findings are the useful part. Two sites can score the same and need completely different work, which is why every check tells you the number it found and what to change rather than only a colour.
What is not in here?
Everything that needs a human. Whether your page says the thing your best customers say about you, whether the order of your sections matches the order of a buyer's questions, where in a paragraph someone gives up. A script cannot read any of that, and that is most of what a real audit is.
The fourteen checks, explained
The tool tells you which checks your page failed. What it cannot tell you is why each check exists, what the failure costs the person reading your page, or how to run all fourteen by hand on a site you cannot paste into a box.
That is what The UX Psychology Audit Checklist is: the same fourteen checks with the reasoning behind each, the three that almost every site fails and why they are missed, and what to fix first when everything is wrong.
Nothing here is held back from the tool. Run it as many times as you like, on any address, without buying anything.