Scores and grades
What the number means, how it is built, and what it cannot tell you.
The score is a number from 0 to 100 with a letter beside it. It is a summary, and like every summary it throws away the detail that tells you what to do. Read it as a starting point, then read the findings.
The grade bands
| Grade | Score | Reads as | In practice |
|---|---|---|---|
| A+ | 95 to 100 | Excellent | Nothing meaningful left. Rare and worth keeping. |
| A | 85 to 94 | Very good | No critical problems, a few recommendations outstanding. |
| B+ | 75 to 84 | Good | Solid page with a handful of real improvements available. |
| B | 65 to 74 | Fair | Working, but leaving something on the table. |
| C | 50 to 64 | Needs work | Usually one or two critical findings dragging it down. |
| D | 35 to 49 | Poor | Several things are actively wrong. |
| F | Below 35 | Critical issues | Something fundamental: often the page cannot be indexed at all. |
How the number is built
Each of the five categories is scored on its own, out of 100, from the checks belonging to it. A failed check costs according to how much it matters: a missing title costs a great deal, a legacy image format costs a little. Passing checks and skipped checks cost nothing.
The overall score combines the five category scores. They do not count equally: indexability counts for the most and schema and media for the least, in the order set out in What we measure.
How much each category contributes is part of how we score and is not published. A formula somebody can read off is a formula somebody can optimise against instead of fixing the page, which is exactly the outcome a score is supposed to prevent. Every finding tells you its own impact, which is the number that actually helps you choose what to do.
What the score is not
| Not a ranking prediction | Where a page ranks depends on every competing page and on intent we cannot see. A page at 95 can sit below one at 60 that answers the query better. |
| Not comparable across page types | A product page and a long article are judged by the same checks but are not the same job. Compare a page with itself over time, and with its own competitors, rather than with the rest of your site. |
| Not a target in itself | Chasing the last five points on a page that already has no critical findings is usually worth less than running the analysis on the next page you have not looked at. |
Moving the number
The fastest way up is always the critical findings, because those cost the most and there are rarely many. Clearing every critical finding on a page scoring in the fifties usually moves it into the seventies on its own.
After that it flattens. Going from 85 to 95 means clearing a long tail of recommendations, each worth a point or two. That is real work with real value, but it is not where you start.
With an account, every analysis of a page is kept, so the score is plotted over time and each run can be compared against the one before it. That comparison is more useful than the absolute number. See History.