By the iSuggest.ai Team · Updated for 2026
Your first audit report can feel like a lot at once: a headline score, a letter grade, a wall of checks that either passed or didn't. Most of it is simpler than it looks, and knowing what each part is actually measuring makes the report far more useful than just glancing at the number at the top. Here's how to read one properly.
The score is built from real signals, not a guess
Every number on the report traces back to something the audit actually observed on the page: how the page identifies itself (title, meta description, canonical), its heading structure, its internal and external links, its images and their alt text, performance signals, and structured data. Nothing is estimated. If a signal can't be measured honestly — say, a property that would require guessing rather than reading what's actually there — it's left out of the calculation rather than scored as if it were checked and failed.
That's also why the score is broken into pillars instead of one flat number: page health carries the most weight, followed by site-wide technical health, then consistency and coverage across pages you've audited, then GEO and entity readiness (the signals that matter specifically for how AI systems read and cite a page). Each pillar is scored out of 100 internally, then weighted into the final grade, so you can see not just that a page needs work but which category the work belongs to.
Page health isn't a simple average
If you audit several pages on the same site, the page health figure isn't just the mean of the individual scores. A tenth-percentile term is folded in specifically so nine strong pages can't quietly hide one badly broken one — a site with a single disastrous page will see that reflected in its overall number rather than diluted away. The home page is weighted twice in this calculation, since it's usually the page an AI crawler or a first-time visitor reaches first.
Some results are refusals, not failures
Not every report ends in a grade, and that's deliberate. Audit fewer than three pages on a site and you'll get a provisional read with no letter grade — there simply isn't enough coverage yet to say something meaningful. If the home page itself can't be reached, the report returns a straightforward F with one clear message rather than a breakdown across pillars that would imply more precision than the situation allows.
There are also hard ceilings that cap a score regardless of how well everything else performs: a home page that isn't indexable caps the report at 49, and a robots.txt file that blocks every major AI crawler caps it at 59. These exist because no amount of on-page polish matters if the page in question can't be found or read in the first place — that's the whole premise of GEO auditing.
What "coverage" actually checks
When a report covers more than one page, coverage is judged against a fixed sample rather than your entire sitemap, so a large site with thousands of URLs isn't punished just for being large. If your sitemap happens to be an index of other sitemaps rather than a direct page listing, don't be surprised that the page count shows as unavailable — that's a correct read of what an index actually contains, not a bug.
Findings are grouped by rule, not repeated per page
If twelve pages are all missing a meta description, the combined report lists that once, as a single finding with all twelve pages attached, rather than repeating the same issue twelve times. It's a small thing, but it's the difference between a report you can act on in five minutes and one that reads like noise. Passing checks show up too, deliberately — a report that only lists faults reads as if nothing on the page works, which is rarely true and isn't the point of an audit.
From report to next step
Reading the report is the first half. The second half is deciding what to do with it — fix the specific findings, re-audit to confirm the fix, and, once a page scores well enough, publish a frozen snapshot of it into an AI directory so Gemini, ChatGPT and Claude have a clean, structured version to cite. The how it works page walks through that full path, and our piece on going from audit to publish covers the workflow in more detail than we have room for here.
Auditing itself costs nothing and has no limit — the credits you'll see mentioned on the account page are only spent when you choose to publish a snapshot, not when you run a report. If that distinction isn't obvious yet, our explainer on the credit system lays out exactly where the free part ends and the paid part begins.
None of this requires memorizing the mechanics above before your first audit. But the next time a score comes back lower than you expected, or a check shows a dash instead of a number, you'll know it's not a mystery — it's a specific, honest read of something real on the page, and usually a specific, fixable one too.