Is the score real? Here is exactly how it works

We read your page the way an AI assistant does, then give it four scores from 0 to 100: Sources, Answers, Structure and Quotability. Your overall score is the plain average of the four.

The first three are checked directly on your page: we count your real sources, look at how long each section is, and read the outline of your headings. The fourth comes from an AI trained to judge how quotable your words are. It only judges what is visible on your page, and it never invents things it can't see.

Example (made up, to show the maths): Sources 60, Answers 90, Structure 75, Quotability 55. Add them up and divide by four: your overall score is 70.

If the AI part is unavailable for a moment, we average the other three and say so. We never fill the gap with an invented number.

Every point comes with a reason you can read.

We check this separately and give it a grade from A to F. First, are the programs that ChatGPT, Claude, Perplexity and Google's Gemini send to read websites allowed onto your site, or does a setting turn them away? Second, can your text be read without the page running scripts, and does the page have clear headings, an author and date, and a sitemap, which is a list of your pages that AI can follow?

Every gap comes with a plain fix, shown from the biggest gain to the smallest. Where it helps, we write the pieces for you to paste into your site, such as the lines that let those AI programs in.

This removes the common reasons an AI can't reach or read your page. It can't make any AI list or quote you, and we check these four, not every AI.

Both parts are real, and we tell you which is which. Three of the four scores are plain checks on your page, with no guessing. The fourth is an AI's judgement, using a fixed 0 to 100 scale, from 'mostly promotional' to 'authoritative, dated, sourced and highly quotable'.

Every score comes with the reasons behind it, marked as a strength, a missing signal or an unsupported claim.

Because you can check our work. We open your outside links to see whether they still work. We show the reason behind every point. And before we promise a gain from a rewrite, we put it into your real text and score the page again.

If a rewrite doesn't clearly help, we don't show it to you.

We can't promise that ChatGPT, Perplexity or any other AI will quote you. Nobody can: those systems change all the time and don't publish how they choose.

Your score is a well-reasoned estimate of how ready your page is, not a live measurement of what an AI says today. Our scoring rules are sensible starting points that we keep tuning as we learn from real pages.

Yes, for the same page. If your page hasn't changed, we reuse the earlier result, so the number doesn't wobble from one check to the next. When you change the page, the score changes with it.

Sources: do you link to real, working pages that back up what you say? A page with no sources scores low, each working source helps, and broken links count against you.

Answers: AI prefers short, self-contained sections it can quote. Sections that are too thin or too long both score lower. Question-and-answer pages get their own, shorter range.

Structure: clear headings in a sensible order give an AI an outline. Jumping from a main heading straight to a small one reads as a broken outline.

Quotability: an AI reads your wording and estimates how likely it is to quote you. It looks for a clear definition up top, direct answers, named sources and dates, and a named author.

Before we tell you a rewrite will help, we swap it into your real page text and score the page again. The gain you see is the measured difference, not a guess.

We only do this when the new text matches your page exactly, and only show it when the improvement is clear. A rewrite that doesn't move the score is left out.

We visit your outside links and see whether each one still opens. For safety we check a limited number per page and never visit private or internal addresses. Links beyond that limit are left unchecked and never counted against you.

Start with the Starting Plan to check your own pages and get a to-do list. Choose Growth to prove your progress with rewrite tests and track rivals. Agencies can offer reports under their own name. Details and prices are on the pricing page.

Send it through the support page and our team will reply. If you'd rather have us do the fixing, you can request a done-for-you project from any plan.