SEO audit example: a real report for our own website, score 95/100

Short answer

This SEO audit example is a real report, not a mock-up. On 18 November 2026 our free SEO audit checked 25 pages of ai-seo-company.pl against 49 rules and scored it 95 out of 100: 0 errors and 3 warnings, all about mobile speed. Below are every number, each issue with its fix, and the format a report should follow.

Run the free audit on your site

www.ai-seo-company.pl · 18 November 2026, 23:12 (Warsaw time)

95

SEO score out of 10025 pages, 49 rules: 0 errors, 3 warnings, 0 notices.

  • Technical SEO100
  • Content100
  • Performance77
  • AI readiness100

The report at a glance

We ran our free SEO audit once on our own domain, on 18 November 2026, 23:12 (Warsaw time), the way any visitor does it: no account and no e-mail. Every number below is copied from the tool’s response, including the ones that do not flatter us.

The audit of www.ai-seo-company.pl in numbers

ItemResult
Domain auditedwww.ai-seo-company.pl
Run18 November 2026, 23:12 (Warsaw time), result after about 26 seconds
Pages checked25
Rules applied49
SEO score95 out of 100
Issues found3: 0 errors, 3 warnings, 0 notices
Previous audit of this domain94 out of 100 on 14 November 2026 (+1 now)
PageSpeed Insights (mobile)performance 83, SEO 100, accessibility 97, best practices 100
Free SEO audit report for www.ai-seo-company.pl: score 95 out of 100, technical 100, content 100, performance 77, AI readiness 100, and three performance warnings with fix instructions
The report as our free audit tool displays it, rendered from the saved result of the 18 November 2026 run for www.ai-seo-company.pl, without the site header. It is the part a visitor sees without giving an e-mail. Here that is the whole list, because the audit found only three issues. When it finds more, the first three are shown and the rest, with the affected page URLs and an e-mailed report, are unlocked with an e-mail address.

Scores by category

The audit sorts its 49 rules into four categories and weights them: technical SEO 35%, content 30%, performance 20% and AI readiness 15%. Three categories scored 100. Performance scored 77, and that is the whole gap to a perfect score.

Category scores in the audit of 18 November 2026

CategoryWeightRulesScoreAgainst 14 NovIssues
Technical SEO35%21100no changenone
Content30%12100+3none
Performance20%1077+23 warnings
AI readiness15%6100no changenone

Weights and rule counts: the free SEO audit’s scoring, as described on its page.

The arithmetic is open. Overall: 0.35 × 100 + 0.30 × 100 + 0.20 × 77 + 0.15 × 100 = 95.4, shown as 95. Performance is 70% the mobile PageSpeed score (83) and 30% the rules. Each of the three warnings costs 12 points of the rules part (4 points × a multiplier of 3), so the rules part is 64, and 0.7 × 83 + 0.3 × 64 = 77.3, shown as 77.

The three issues and how to fix them

The audit reported no errors and no notices. All three warnings are about the speed of the home page on a phone, and they are connected: heavy images slow the largest element, and the largest element drags down the PageSpeed score. For each issue we show what the rule checks, what was measured and the fix the tool gives.

Warning · Performance

Images to optimise (over 150 KB to save)

What the rule checks
Whether the images on the home page could be much smaller. The rule reports a problem when PageSpeed Insights estimates more than 150 KB of possible savings, because old formats, files larger than the space they fill and weak compression all slow the first screen down.
What we measured
Possible savings: 193 KB. The three biggest single savings: 50 KB, 49 KB (feature.webp), 35 KB. A file that is already WebP, like feature.webp, can still be flagged when it is larger than its slot or compressed too lightly.
How to fix it
Save images as WebP or AVIF, at their display size (srcset for different screens) and with about 75-85% compression.

Warning · Performance

Mobile performance needs improvement (PageSpeed 50-89)

What the rule checks
The mobile performance score from PageSpeed Insights. Lighthouse calls 50 to 89 “Needs Improvement” and 90 to 100 “Good”.
What we measured
83 out of 100: the median of three runs (86, 83, 76). The runs differ by 10 points, which is why the audit takes three and does not trust a single one.
How to fix it
Start with the largest element (LCP): a hero image in WebP with preload and without lazy loading.

Warning · Performance

The largest element loads too slowly (LCP over 2.5 s)

What the rule checks
Largest Contentful Paint, the time it takes for the main content to appear and one of Google’s Core Web Vitals. Google treats 2.5 seconds or less as good and more than 4 seconds as poor, measured at the 75th percentile of page loads.
What we measured
3.2 s in this run, so in the “needs improvement” band between 2.5 and 4 seconds. Google applies its thresholds to many page loads, so one measurement is a pointer, not a verdict.
How to fix it
Speed up the image or heading on the first screen: a smaller file, preload, a faster server.

LCP is one of Google’s Core Web Vitals, so the third issue is the one to check against Google’s own thresholds first. Our guide to Core Web Vitals and Google rankings explains the thresholds in detail.

Rules passed and failed, by category

In this run 3 of 49 rules reported a problem, all in performance. The other 46 reported nothing: their check passed or did not apply to this site. Open a group to see its rules, grouped as the tool groups them.

Technical: 21 rulesall 21 reported nothing
  1. Pages return an error (4xx/5xx)Error, 10 pts
  2. Redirect chainsWarning, 3 pts
  3. The site does not run on HTTPSError, 10 pts
  4. The http:// version does not redirect to https://Warning, 4 pts
  5. The www and non-www versions do not lead to a single addressWarning, 3 pts
  6. No robots.txt fileNotice, 1 pt
  7. robots.txt blocks the entire siteError, 15 pts
  8. No sitemap (sitemap.xml)Warning, 4 pts
  9. No canonical URLWarning, 3 pts
  10. Pages excluded from the index (noindex)Notice, 2 pts
  11. Home page excluded from the index (noindex)Error, 15 pts
  12. The sitemap lists URLs that should not be indexedWarning, 3 pts
  13. Missing lang attribute in <html>Warning, 2 pts
  14. The page is not mobile-friendly (no meta viewport)Error, 6 pts
  15. Resources loaded over http:// on an https:// pageWarning, 2 pts
  16. Non-existent addresses return a page with code 200 (soft 404)Error, 5 pts
  17. Language versions without hreflang tagsWarning, 4 pts
  18. Errors in hreflang tagsWarning, 3 pts
  19. Pages without hreflang on a multilingual siteNotice, 1 pt
  20. Broken internal linksError, 6 pts
  21. Broken external linksNotice, 1 pt
Content: 12 rulesall 12 reported nothing
  1. Missing page title (title)Error, 8 pts
  2. Title too long (over 60 characters)Warning, 2 pts
  3. Title too short (under 20 characters)Notice, 1 pt
  4. Duplicate titles across several pagesWarning, 4 pts
  5. Missing meta descriptionWarning, 3 pts
  6. Meta description too long (over 160 characters)Notice, 1 pt
  7. Duplicate meta descriptionsNotice, 2 pts
  8. Missing H1 headingError, 6 pts
  9. Several H1 headings on the pageNotice, 1 pt
  10. Little text in the HTML (under 250 words)Warning, 4 pts
  11. Images without alternative text (alt)Warning, 2 pts
  12. No Open Graph tagsNotice, 1 pt
Performance: 10 rules3 reported a problem, 7 nothing
  1. Very heavy home page (over 5 MB)Error, 6 pts
  2. Heavy home page (over 3 MB)Warning, 3 pts
  3. Images to optimise (over 150 KB to save)Warning, 4 pts, reported a problem
  4. Images without dimensions (width and height)Notice, 2 pts
  5. Slow server response (over 1.5 s)Warning, 4 pts
  6. Large HTML document (over 500 KB)Notice, 2 pts
  7. Low mobile performance (PageSpeed below 50)Error, 8 pts
  8. Mobile performance needs improvement (PageSpeed 50-89)Warning, 4 pts, reported a problem
  9. The largest element loads too slowly (LCP over 2.5 s)Warning, 4 pts, reported a problem
  10. Elements jump around while loading (CLS over 0.1)Warning, 3 pts
AI readiness: 6 rulesall 6 reported nothing
  1. Content invisible without JavaScript (under 50 words in the HTML)Error, 8 pts
  2. robots.txt blocks AI botsWarning, 8 pts
  3. No company structured data (Organization / LocalBusiness)Warning, 5 pts
  4. Invalid structured data (JSON-LD)Warning, 3 pts
  5. llms.txt file without links to pagesWarning, 2 pts
  6. No llms.txt fileNotice, 1 pt

SEO audit format: what a report should contain

A good audit report lets someone who was not in the room decide what to do next. If you are writing an audit yourself, this is the SEO audit format we follow, and the right-hand column shows how this example fares. Use our SEO checklist template for the checks you fill it with, or the technical SEO audit checklist for the technical part.

The parts of an SEO audit report and how this example covers them

Part of the reportWhat it should containIn this example
Summary and scoreOne overall number, a score per category and the count of issues by severity.Yes: 95/100, four category scores, 0 errors, 3 warnings, 0 notices.
Scope and methodThe domain, the date, how many pages were checked and which tools and data sources were used, so a reader knows what the result covers.Yes: 25 pages, 49 rules, our AISEOCompanyAudit crawler and Google PageSpeed Insights, date and time of the run.
Findings by severityErrors first, then warnings, then notices, each with a plain-language title.Yes: three warnings, no errors, no notices.
EvidenceThe measured value or the affected URLs behind each finding.Measured values (193 KB, 83/100, 3.2 s). Affected URLs: not applicable, because all three rules concern the home page as a whole.
Fix instructionsWhat to change, in words a developer can act on.Yes: one instruction per issue.
PrioritiesThe order of work, with the biggest effect first.Severity decides the order. With only warnings, start with the one that moves the score most: mobile performance.
Change since the last auditA comparison with the previous run, overall and per category.Yes: +1 against 14 November 2026 (94/100), with a change per category.
LimitsWhat the audit did not check, stated plainly.Yes: see “What an automated audit cannot do” below.

How to read a report like this

  1. Read the category scores before the total. A 95 hides a 77. The total is a weighted average, so one weak category is softened by three strong ones. Here the only category below 100 is performance.
  2. Separate errors, warnings and notices. In this tool an error rule is worth 5 to 15 points, a warning 2 to 8 and a notice 1 or 2, before a multiplier of 3. One failed error rule can cost a category up to 45 points, which is why errors come first.
  3. Check whether a problem is site-wide or per page. A site-wide rule, such as the three above, costs its full penalty. For a per-page rule, one affected page already costs 40% of the penalty, and the cost grows with the share of pages affected.
  4. Compare measured values with Google’s thresholds. An LCP of 3.2 s is not good against Google’s 2.5 s, and a PageSpeed score of 83 is in the “Needs Improvement” band, below the 90 that Lighthouse calls “Good”.
  5. Compare with the previous audit and allow for noise. The score moved from 94 to 95 in four days. Speed measurements vary between runs (86, 83 and 76 here), so treat a change of a few points in performance as noise unless it repeats.
  6. Fix in order, then run the audit again. Fix errors first, then the warnings with the biggest effect, then run the audit again to confirm the change in the score.

What an automated audit cannot do

The free audit is fast and consistent, and that comes with limits. These are the ones that matter when you read the score of 95 above.

  • It checks a sample of up to 25 pages: the home page, the links on it and URLs from the sitemap. On a large website most pages stay outside the audit.
  • It measures speed in PageSpeed Insights on mobile, for the home page only. Other templates, such as long articles or product pages, can behave differently.
  • It reads the HTML as the server sends it, without running JavaScript, the way most AI bots do. Content added by scripts may be invisible to it.
  • It looks for duplicate titles and descriptions only among the pages it checked, and it checks external links only for 404 and 410 codes.
  • It does not check links pointing to your site, keywords, Google rankings or Search Console data.
  • It cannot judge whether a page answers what a searcher wants, whether the content is accurate and useful, or how you compare with named competitors. A person has to read the pages for that.

A manual audit covers the whole website and adds keywords, competitors, the link profile and Search Console data, with a plan of changes. Our SEO audit service does that, starts from €350 net and takes 3 to 5 business days; the current prices of all packages are on the SEO pricing page.

See your own report

Enter your website address and get the same kind of report in 20 to 90 seconds: a 1-100 score, four category scores and the top issues with fixes. No sign-up; up to 5 audits per day from one IP address.

Run the free SEO audit

Frequently asked questions

A summary with a score, the scope and method, findings ordered by severity, the evidence behind each finding, fix instructions, priorities, a comparison with the previous audit and the limits of the check. The table above shows each part in this real report.

Yes. It is the result of one run of our free SEO audit on our own website on 18 November 2026, 23:12 Warsaw time. The numbers are copied from the tool’s response, including the three warnings and the 77 in performance. You can run the same audit on your site and compare.

In this tool, 80 or more means the basics are fine and it is worth turning to content and links. From 50 to 79 there are problems that limit visibility, so start with errors, then warnings. Below 50 there are technical gaps that can stop Google from indexing the site. Scores from other tools use other formulas and are not comparable.

No. It means that 46 of 49 automated rules reported nothing on the 25 pages checked. The audit does not look at backlinks, rankings, search intent or content quality, and our own report still has three performance warnings to fix.

This page is not a download: it shows a finished report and the format a report should follow. For the checks to fill it with, use our SEO checklist template, a free PDF with 50 points.

Sources

  1. web.dev: Largest Contentful Paint (LCP), thresholds of 2.5 and 4.0 seconds at the 75th percentile (accessed November 18, 2026)
  2. Chrome for Developers: Lighthouse performance scoring (0-49 Poor, 50-89 Needs Improvement, 90-100 Good) (accessed November 18, 2026)
Start your project →