# SEO Health Score Explained: What It Measures and Misses > What an SEO health score measures, how Semrush, Ahrefs and Lighthouse calculate theirs, what counts as a good score, and why Google never sees the number. URL: https://seorecheck.com/blog/seo-health-score-explained · Published: 2026-10-25 · [Українська](https://seorecheck.com/ua/blog/seo-health-score-explained.md), [Español](https://seorecheck.com/es/blog/seo-health-score-explained.md) An **SEO health score** is a 0–100 number that an audit tool gives your website after its crawler checks your pages against its own list of rules. It shows how many technical issues the tool found and how serious it thinks they are. Every vendor calculates it differently, Google doesn't use it, and a high score doesn't guarantee rankings. That doesn't make the score useless. It's a quick summary of your site's technical condition, and it works well for comparing your site with itself over time. The problems start when people treat it as a ranking metric, compare scores across tools, or fix whatever moves the number fastest. This guide covers how the main tools calculate the score, what a "good" score means, and how to use it without being misled. ## What is an SEO health score? An SEO health score sums up an automated site audit in one figure. A crawler visits your pages, runs a fixed set of checks ([broken links](https://seorecheck.com/blog/404-errors-seo-impact), [redirect chains](https://seorecheck.com/blog/301-vs-302-redirects-seo), missing titles, noindex tags, slow pages, invalid markup), sorts each finding by severity and turns the result into a percentage or a 0–100 value. Tools give it different names: "Site Health" in Semrush, "Health Score" in Ahrefs, "SEO score" in Lighthouse and in many free online checkers. They all work the same way. **The score measures how well your site follows that particular tool's checklist.** It doesn't measure traffic, rankings, content quality, links or whether people searching your topic find your pages useful. So the score answers one narrow question: "How many known technical problems did this crawler find, weighted by how serious this tool thinks they are?" That's a useful question. It just isn't "How good is my SEO?" ## How do SEO tools calculate the health score? Each tool uses its own formula, so the same site can score very differently in two tools on the same day. Some tools count pages with errors. Others weight issue types. Lighthouse averages pass/fail audits for one page. Here's how the documented formulas compare. | Tool | What the score is based on | Issue types that lower it | Scope | |---|---|---|---| | Ahrefs Site Audit | Share of internal URLs with no error-level issues | Errors only. Warnings and notices don't count | Every crawled internal URL | | Semrush Site Audit | Errors and warnings relative to checks performed, weighted by check type | Errors weigh more than warnings and notices | Crawled pages, independent of page count | | Lighthouse SEO category | Average of individual SEO audits | Failed audits, equally weighted | One page, one load | | Lighthouse Performance | Weighted average of lab metrics | Slow metric values | One page, one load | ### Ahrefs: the share of error-free pages The [Ahrefs help centre](https://help.ahrefs.com/en/articles/1424673-what-is-health-score-and-how-is-it-calculated-in-ahrefs-site-audit) explains that the Health Score is the number of internal URLs without error issues, divided by the total number of internal URLs, times 100. Only "Error" issues lower the score. Warnings and notices don't affect it. Ahrefs labels 0–30 as Weak, 31–70 as Fair, 71–90 as Good and 91–100 as Excellent. Here's an example. A 200-page site where 40 product pages share one error-level issue scores (200 − 40) / 200 × 100 = 80%. Fix that one template problem and all 40 pages become error-free at once. ### Semrush: weighted by check type [Semrush's documentation](https://www.semrush.com/kb/114-total-score) says Site Health is based on the errors and warnings found on crawled pages, and that errors count more than warnings. The calculation takes into account the type of check, the number of unique checks triggered and the number of issues fixed for each check. Semrush notes that the score doesn't depend on how many pages were crawled. Fixing every instance of one check raises it more than fixing two errors from two different checks. ### Lighthouse: one page, equal weights Lighthouse scores a single URL. According to [Chrome's documentation on the meta description audit](https://developer.chrome.com/docs/lighthouse/seo/meta-description), each [SEO audit](https://seorecheck.com/blog/what-is-an-seo-audit) carries equal weight in the Lighthouse SEO score, except the manual "Structured data is valid" audit. The Performance score works differently. [Chrome's performance scoring guide](https://developer.chrome.com/docs/lighthouse/performance/performance-scoring) gives the Lighthouse 10 weights: Total Blocking Time 30%, Largest Contentful Paint 25%, Cumulative Layout Shift 25%, First Contentful Paint 10% and Speed Index 10%. Scores of 90–100 are green (good), 50–89 orange and 0–49 red. ## What is a good SEO health score? There's no universal "good" score, because each scale measures something different. As a rough guide, 90+ in a crawler-based tool means few error-level issues remain, and 70–90 is typical for an active site with some backlog. What matters more is which issues sit behind the number, which pages they affect and whether the score is going up between audits. These two sites show why the number alone doesn't settle it: - **Site A scores 72%.** Hundreds of old tag archive pages have duplicate titles. That's annoying but low impact: they get no traffic and could be noindexed or removed. - **Site B scores 97%.** The crawler found an error on only a handful of URLs. One of them is the main service page, which a plugin update accidentally set to `noindex`. Site B has the higher score and the more urgent problem. In Ahrefs' formula, that noindexed page counts the same as any other URL with an error. In your business, it may be the page that brings in most of your leads. The score doesn't know which pages earn money. You do. So a good score is one that **has no critical issues on important pages behind it**, whatever the exact figure. ## Does Google use an SEO health score? No. Google doesn't see or use any third-party health score. The number exists only inside the tool that produced it. Google's [SEO starter guide](https://developers.google.com/search/docs/fundamentals/seo-starter-guide) states that "there are no secrets here that'll automatically rank your site first in Google". No checklist score can promise rankings. Some common audit checks even cover things Google says don't matter much. The same starter guide says: - Google Search doesn't use the keywords meta tag. - Content length alone doesn't matter for ranking. - From Google's point of view, it doesn't matter if headings are used out of order. - Content available under multiple URLs is "fine; don't fret about it". Consolidating duplicates helps, but it's not a penalty. If a tool takes points off for a missing meta keywords tag or a skipped H3, your score drops even though nothing that affects Google has changed. That's why you should read the findings and not just the total. ## Why did my SEO health score change when nothing broke? A score can move without any real change on your site, and it can stay flat while something important breaks. The usual reasons: 1. **Different crawl scope.** More or fewer URLs crawled (a new crawl limit, a sitemap change, faceted URLs found for the first time) changes the denominator in page-based formulas. 2. **New or updated checks.** Vendors add rules. A site with no changes can drop overnight because the checklist got longer. 3. **One heavy issue offsets many light ones.** [Semrush warns](https://www.semrush.com/kb/114-total-score) that your total issue count can go down while your score also goes down, if a new issue with a strong effect appears while you clean up less significant ones. 4. **Lab variability.** [Chrome's documentation](https://developer.chrome.com/docs/lighthouse/performance/performance-scoring) says Performance scores fluctuate because of A/B tests, changes in traffic routing, different devices, browser extensions and antivirus software. 5. **Configuration changes.** Ignored issue types, a different user agent or JavaScript rendering switched on or off produce different findings. To compare audits fairly, keep the crawl settings the same and compare **issue by issue**, not total against total. Our guide to an [SEO audit before and after comparison](https://seorecheck.com/blog/seo-audit-before-after-comparison) shows how to sort every finding into fixed, improved, unresolved and new. ## What does an SEO health score not tell you? A health score is built from what a crawler can detect automatically. Whole areas of SEO are outside its reach: - **Content quality and search intent.** Whether a page actually answers the query better than the competing results. - **Links and reputation.** Who links to you and whether people trust your brand. - **Real-user performance.** Lighthouse is a lab test. Google's [Core Web Vitals](https://seorecheck.com/blog/core-web-vitals-guide) use field data from real visitors at the 75th percentile, which can differ a lot from one lab run. - **Business value.** Which pages bring in revenue, so which issues are actually urgent. - **Indexing reality.** A crawler sees what's on your pages. Only Search Console shows what Google actually indexed and why it left other pages out. - **AI search visibility.** Most scores don't cover AI crawler access in robots.txt, content that only appears after JavaScript runs, or entity markup. Each needs its own checks. ## How should you use an SEO health score? Use the score as a trend line and a way into the findings, not as a goal in itself. A simple routine: 1. **Record a baseline** with fixed crawl settings, and note the date and tool version. 2. **Open the findings, not the summary.** Sort them by severity and by the pages affected. 3. **Filter to your important pages** first: home page, main service or category pages, top landing pages from Search Console. 4. **Fix template-level issues before one-off ones.** One fix to a template can clear hundreds of URLs, which is also why it moves page-based scores the most. 5. **Skip findings that don't affect search.** Meta keywords and heading order are examples Google itself names. 6. **Re-audit with the same settings** after the fixes are live, and check that critical issues are gone. Don't just look at the number. For a full method to order a long list of issues by impact, scope and effort, see [how to prioritize SEO fixes](https://seorecheck.com/blog/how-to-prioritize-seo-fixes). If you're testing free checkers, our guide to a [website SEO checker online](https://seorecheck.com/blog/website-seo-checker-online) explains what those quick scores cover and what they leave out. ### How SEORecheck reports its score The SEORecheck preview shows an SEO score next to real issues with example URLs from your pages, so you see the evidence behind the number right away. You can see the format in the [sample report](https://seorecheck.com/sample). The full report lists every finding with the affected URLs and fixes, plus a 30/60/90-day plan (see the [full sample](https://seorecheck.com/sample/full)). The AI search chapter is reported as checks passed, not as a separate "GEO score", because a single number would hide which check failed. A recheck 30–60 days later compares every issue with the previous audit ([comparison sample](https://seorecheck.com/sample/compare)), so progress is measured by fixed issues, not by score movement. Prices vary by country and are listed on the [pricing page](https://seorecheck.com/pricing). If you want to see which issues sit behind your own number, a [free audit preview](https://seorecheck.com/request) shows your score along with real crawl, indexing and performance findings from your pages. ## Key takeaways - An SEO health score is a tool-specific summary of automated checks, not a Google metric. - Ahrefs counts the share of internal URLs without errors. Semrush weights errors and warnings by check type. Lighthouse averages equally weighted audits for one page. - Scores from different tools can't be compared. Compare one tool with itself, using the same crawl settings. - A high score can hide a critical issue on your most valuable page, and a low score can come from harmless low-value URLs. - Measure progress issue by issue (fixed, improved, unresolved, new), not by the change in the total. --- ## Want to know what to fix on your site? Get a free SEO audit preview: your score and real issues, with evidence from your pages. - [Get a free SEO audit preview](https://seorecheck.com/request) - [See a sample report](https://seorecheck.com/sample)