How to Read Your SEO Audit Results and What to Fix First?

The DoSieci SEO Check result is a number from 0 to 100 with a letter grade, plus a list of checks split into six categories. The number itself isn't the point — it's there so you can see within a few minutes whether a problem is serious or cosmetic, and where to start.

What the score number actually means

The scoring is deterministic: each check carries a fixed weight, and the score is the sum of points earned across six categories. Accessibility and indexability carry 30 points, metadata 20, content 20, and links and images, structured data, and mobile-friendliness and performance carry 10 each. That split isn't arbitrary: a page with perfect titles that Google can't index shouldn't get a high score.

Alongside the score, the report shows coverage as a percentage. If some data couldn't be collected — for example, a performance measurement or real-user data — coverage drops, but the missing checks don't lower the score. So low coverage means “partial report,” not “your site has a problem.”

This is a DoSieci score calculated from the publicly available data for one URL. It is not a Google ranking score and doesn't tell you your position in search results.

Start with blockers, not cosmetics

The report puts the three most important problems at the top precisely because order matters. First check whether the page can even reach the search results at all:

• a meta robots tag or X-Robots-Tag header with noindex — the most common silent error. A block left over from a draft version stays in place after publishing, and no other fix will work until you remove it;
• a canonical tag pointing to a different address — this tells Google to index a different page instead. After a migration or copying a template, it's common for every subpage to point at the homepage;
• a block in robots.txt — a Disallow rule that covers the address you're trying to promote;
• a response code other than 200, or a redirect chain — the page may actually be responding at a different address than you think.

Only once the blockers are gone does it make sense to work on the title, headings, and content. Doing it in the reverse order just wastes time.

How to read each category

Metadata and document. The title is the headline of your result in the search engine. A default “Home Page” title, the same title repeated across many subpages, or a title unrelated to the content all cost you real clicks. The same goes for having one clear H1 heading — not because Google “requires a single H1,” but because a page without a clear topic answers the query poorly.

Content and intent. The test checks how much visible content is present in the HTML code and whether the first paragraph states what the page is about. If the content is only added by JavaScript, the report will show it — Google will see it with a delay, and some bots won't see it at all.

Links and images. Empty link text (“click here”), HTTP links on an HTTPS page, missing ALT descriptions, and images without set dimensions. The last of these is responsible for layout shifts while the page loads.

Structured data. JSON-LD syntax errors and schema that describes something other than the visible content. Correct structured data doesn't boost rankings by itself — it organizes what the search engine can show alongside your result.

Mobile-friendliness and performance. Missing meta viewport, zoom disabled, no compression, and — where data is available — a performance score and Core Web Vitals. This only covers the basics; a full speed diagnosis needs measurements and access to the site.

What this score doesn't tell you

• Whether the page is in Google's index. Indexability tells you whether Google can index the address. Whether it's actually in the index is something you check in Google Search Console.
• What rankings you have, and for which keywords. The test analyzes one URL, not search engine data.
• How the whole site performs. Only the exact address you submit is analyzed, plus robots.txt and the sitemap. Duplicate content, keyword cannibalization, and site architecture require crawling the entire site.
• Whether the content is better than your competitors'. The checks are technical and formal; they don't judge the substantive quality of the text.

A sensible order of action

1. Remove indexing blockers and fix the canonical tag.
2. Clean up the title and description of the page you want to promote — this is the fastest visible change.
3. Fill in content wherever the report shows the page doesn't clearly state what it offers.
4. Fix internal linking and image descriptions.
5. Only at the end, tackle performance and structured data — once the first four points are done.

After making changes, run the test again on the same address. The report has a permanent link, so you can compare what actually changed.

When it's worth asking for a review

If the report shows a problem you can't clearly pin down to a spot in the code or the configuration, send us the link to your result through the contact form and describe what you're trying to achieve. A high score doesn't rule out asking, either — the test checks form, not whether the page actually answers your customers' real questions.