The Final Score

The Final Score is the single number at the top of the visibility panel. It answers a broader question than the AI Score does: how well is this brand positioned to be recommended by AI assistants — which depends both on what assistants say today and on whether your own site can be read by the crawlers that feed them.

Final Score = 0.8 × average AI Score + 0.2 × SEO Score Both terms are 0–100, so the result is 0–100. Rounded to one decimal place.

The two halves

80% · measured outside Average AI Score The mean of every scorable probe in the current scope — all intents, all engines, all regions unless you have filtered. This is the part that moves when your visibility changes.

20% · measured on your site SEO Score The mean of four Lighthouse categories from a PageSpeed Insights run of your homepage: Performance, Accessibility, Best Practices and SEO.

Why 80 / 20

AI visibility is the thing being sold and the thing you can move week to week, so it dominates. But a site that AI crawlers can’t load, or that scores badly on the fundamentals every engine’s retrieval pipeline shares, sets a ceiling on that visibility no prompt strategy can lift. The 20% term keeps that ceiling visible on the headline number instead of hiding it in a separate tab.

The weights are fixed and not configurable — a score that means something different per account can’t be compared over time or across brands.

The SEO Score

The SEO half comes from Google PageSpeed Insights, run against the target’s homepage:

Category	What it measures	What it affects
Performance	Load and interactivity — Core Web Vitals (LCP, total blocking time, layout shift)	Bounce rate, conversions, search ranking
Accessibility	Contrast, labels, keyboard navigation, assistive-tech support	Audience reach, WCAG/ADA compliance
Best Practices	HTTPS, console errors, safe dependencies, correct image sizing	Trust, security, maintainability
SEO	Crawlability, meta tags, mobile-friendliness, crawlable links	Organic traffic and discoverability

SEO Score = mean(Performance, Accessibility, Best Practices, SEO) Each category is 0–100. Categories that PageSpeed did not return are left out of the mean rather than counted as zero.

The audit runs automatically alongside a probe run when recent data exists, and can be triggered on demand from the Site Audit tab.

When there is no audit

If no Lighthouse data is available — the audit hasn’t run yet, PageSpeed timed out, or the site is behind bot protection — the SEO term is dropped, not zeroed:

Final Score = average AI Score A missing audit is missing information, not a failing grade.

Zeroing it would cost 20 points of headline score for a measurement that was never taken, which would read as a product regression on a day nothing changed.

Scope: what “average AI Score” covers

The Final Score is recomputed for whatever scope you’re looking at, using the same formula:

Scope	AI half	SEO half
All (default)	Every scorable probe in the run	Site-level SEO Score
A single region	Probes from geo-adjustable engines in that market	Same site-level SEO Score
Global engines	Probes from engines with no location signal	Same site-level SEO Score
A single engine	That engine’s probes	Same site-level SEO Score

Identical SEO halves across regions are correct, not a bug

Your site is the same site in every market, so the SEO Score can’t vary by region. All regional variation in the Final Score comes from the AI half. Region cards label this explicitly with a site-level marker so the repeated number doesn’t look like a data error.

Score history

Each run contributes one point to the score history chart. The stored headline number is always the all-scope Final Score, computed over every probe in the run — including engines with no geographic signal.

That matters when you filter: the per-region score you see on the panel is computed for that scope at read time and can legitimately differ from the headline. The history line stays on one definition so old and new runs remain comparable; changing the definition retroactively would create a break in the chart at the point the change shipped.

Per-region history is only available for runs made after per-region tracking shipped. For earlier runs the chart falls back to the overall trend and says so.

Averages & denominators → Exactly which probes go into "average AI Score". Site Audit → Where the SEO half comes from, and how to move it.