On this page
What the SEO checker reads, and why
The SEO checker fetches one page the way Googlebot does and judges fourteen parts: whether Google may index it, whether it answers over https, the title measured in pixels, the canonical, the H1, the meta description, mobile scaling, internal links and their anchor text, the sitemap, structured data, alt text, language and hreflang, Open Graph and the amount of readable text. You get a score, the first fix, the two after it, and one sentence per part on what to do. No email address, no account.
The order is Google's own. Its Search Essentials list three technical requirements: Googlebot is not blocked, the page answers with status 200, and it has content Google can index. Only once those hold do title, links and alt text count. If a page is missing from Google altogether, start with website not showing on Google. How Belgian small business sites score on these parts is in our study of Belgian business websites in 2026: 36.5 percent lack a meta description. What to fix next, and in which order, is in how to improve SEO.
| Part | What the checker measures | Why it matters | What to do when it is wrong |
|---|---|---|---|
| Access | noindex in meta robots or the X-Robots-Tag, robots.txt for Googlebot | Without access Google does not index the page | Remove the noindex or the robots.txt rule, then request a recrawl in Search Console |
| Reach | Status, redirects, https | Google only indexes pages that answer 200 | Link straight to the final address, redirect http to https |
| Title | Count, characters and width in pixels | The title is what people read in the results | Search term first, around 600 pixels, test it in the SERP preview tool |
| Canonical | Does it point to the page itself | A canonical elsewhere asks Google to index another page | Point it to this exact address |
| H1 | One, none or several | The main heading helps Google pick the title and topic | One H1 with the search term |
| Meta description | Length, nosnippet | Google uses it when it sums up the page better than your text | One description per page, the reason to click first |
| Mobile | Viewport tag | Google indexes the mobile version of your site | Add the viewport tag to the head |
| Internal links | Count, links without text, vague anchor text | Google finds pages through links and reads their text | Say where the link goes, not "read more" |
| Sitemap | In robots.txt or at /sitemap.xml | Helps Google find new pages | Add a Sitemap line to robots.txt |
| Structured data | JSON-LD, and whether it is valid JSON | Can enable rich results, without a guarantee | Organization or LocalBusiness on the homepage |
| Alt text | Images without an alt attribute | For Google Images and screen readers | Describe the image, no keyword lists |
| Language and hreflang | lang, self-reference, x-default, invalid codes | The right language version for the right reader | Every version lists itself and the others, with valid codes |
| Open Graph | og:title and og:image | How a shared link looks on LinkedIn | Both tags on every page |
| Text | Words of readable text | A page without an answer rarely ranks | Answer the searcher's question in full |
The score in four groups
What the score weighs
Access29%
noindex and robots.txt, status and https
Core of the page39%
title in pixels, canonical, H1, meta description, mobile
Coherence25%
internal links, sitemap, structured data, alt text, language and hreflang
Extras7%
Open Graph, amount of text
Two ceilings
A noindex, a robots.txt that blocks Googlebot or an error code keeps the score at 30 or below, however good the rest is. A missing basic (no title, no H1, no internal link, a canonical elsewhere) keeps it at 60 or below.
What no tool sees
- What the searcher really wants to know
- Whether your text answers that better than the rest
- Links from other sites to yours
- Your positions and clicks: those live in Search Console
The share per group follows the weights in the checker: access and the core of the page count most, Open Graph least, because it mostly decides how a shared link looks on LinkedIn or Facebook.
What an SEO score does and does not tell you
The score counts a part in order as a full point and a part to look at as half a point, weighted by how much it matters. Access and reach weigh three times as much as alt text; title, canonical and mobile twice. If something keeps the page out of Google (a noindex, a robots.txt that blocks Googlebot, an error code), the score stays at 30 or below, however good the rest is. A page can never score high while it is invisible. If a basic is missing (no title, no H1, not a single internal link, a canonical pointing elsewhere, no viewport, or plain http), the score stays at 60 or below: a page without a main heading never reads as "the basics are in place". And what does not apply earns no free points: on a page without images, alt text does not count.
| Score | What it means | What to do now |
|---|---|---|
| Up to 30, with a block | Google does not take the page in as it stands | Fix the block and request a recrawl; the rest waits |
| Below 70 | One or more basic faults, such as no title, no H1, no internal link or a canonical elsewhere | Work through the list in the checker's order, an afternoon for a small site |
| 70 to 89 | The basics are in place, with points that cost clicks | Rewrite title and description, give your links real text |
| 90 to 100 | The technical side of the page is in place | Work on content and on links from other sites |
What the score does not tell you is whether you rank. Four things no tool sees from outside: what the searcher really wants, whether your text answers it better than the pages on top now, which other sites link to you, and how Google shows and clicks your pages. The last one lives in Search Console, which since 31 August also reports AI answers (see the Search Console AI performance report).
Google's SEO Starter Guide also names what matters less than many tools suggest. Google does not use the keywords meta tag. There is no minimum or maximum word count: length alone does not matter. The number and order of headings do not matter, and E-E-A-T is not a ranking factor. So this checker gives no points for word count and only flags thin pages under 150 words. Google says changes take effect in anything from a few hours to several months; more on that in how long until my website shows on Google.
After the check: working through the fixes
- Access and reachRemove a noindex, change robots.txt or fix the error code. Then request a recrawl through URL Inspection in Search Console; without it a revisit can take months. A quarter of an hour, with the biggest effect.
- Title and H1Search term at the front of the title, around 600 pixels, and an H1 that says the same in a full sentence. Google picks up a new title within days to weeks.
- Canonical and hreflangA canonical to the page itself. On a multilingual site every version lists itself and the others, with valid codes and an x-default.
- Links with real textReplace "read more" and "click here" with the name of the page the link goes to. Give a linked icon or image a label or alt.
- Check again, then measureRun the checker after each change. After a few weeks, Search Console shows whether impressions and clicks on your search terms move.
Multilingual sites: the part most checkers skip
A company that sells in Belgium, the Netherlands and Germany usually runs one site in three or four languages. That is where hreflang decides which version a searcher gets. Google's rules are strict: every version lists itself and all others, and if two pages do not point to each other, the tags are ignored. The language code comes first (ISO 639-1), the country second and only if you need it (ISO 3166-1). A country alone is not valid, and Google says plainly that EU, UN and UK have no effect. Google also does not use hreflang or the lang attribute to detect a page's language: it reads the text. The checker flags codes Google cannot read, a missing self-reference and a missing x-default.
| Market | Valid hreflang | Common mistake | What goes wrong |
|---|---|---|---|
| Dutch in Belgium | nl-BE | be | be is the code for Belarusian, not Belgium |
| French in Belgium | fr-BE | fr-UK, or fr only on a site with France too | UK has no effect; without a country, Belgian and French readers share one version |
| English in the UK | en-GB | en-UK | Google ignores UK, so the tag does nothing |
| German in Austria and Switzerland | de-AT, de-CH | at, ch | at is no language code; ch is Chamorro |
| Ukrainian | uk | uk used for the United Kingdom | Your English page is announced as Ukrainian |
| Everyone else | x-default | No x-default | Readers outside your markets get a random version |
In Belgium this is daily work: one company, two or three official languages, and customers who search in Dutch in Ghent and in French in Liège. We build such sites in code, with hreflang generated from the same list as the pages, so every page knows its language partners. See multilingual website development; K&A Law is a site of ours in four languages.
Examples by type of business
A British firm entering the Benelux. The new Dutch pages carry hreflang en-UK on the English side and nothing on the Dutch side. Fix the code to en-GB, add nl-NL and nl-BE (or plain nl if one version serves both), and make every page list all of them.
A German manufacturer with three language versions. The canonical of every English page points to the German one. Google then indexes the German page only. Give every version its own self-referencing canonical; hreflang links them.
A Dutch webshop shipping to Belgium. Filter pages (?colour=red) without a canonical and a thousand products with the same description. Start with canonicals on category and product pages, then alt text on your best sellers.
A Belgian services company with ten people. Each office has a page, but they share one title. Give each a title with its city and LocalBusiness details; local visibility also depends on your Google Business Profile, see local SEO.
Doing it yourself? The SEO checklist has twenty checks in order of impact. To see whether ChatGPT and Claude may read your site, use the AI crawler checker; for load times, the website speed test.
Score, limits and privacy
Is this SEO checker free?
Yes. Enter an address and you get your score at once, without an email address or account. Each visitor can run it twenty times an hour and a hundred times a day, so it stays fast for everyone.
My score is high, so why am I not on page one?
The checker tells you whether Google can read your page well. Ranking also depends on how well your text answers the question, who links to you and what competitors do. A score of 90 or more is the condition, not the result.
Why could the checker not read my site?
The checker reads your site as VBSMarketingTools and respects your robots.txt. If that keeps all crawlers out, or bot protection blocks test tools, it stops. Googlebot often notices nothing. Use URL Inspection in Search Console to see what Google sees.
Does it check every language version?
One page at a time: the address you enter. Run it on each language version of the pages that matter. It does check that the page lists its other versions correctly in hreflang.
How fast will I see the effect of a fix?
Google says anything from a few hours to several months. A new title is usually picked up within days to weeks. A removed noindex goes faster if you request a recrawl in Search Console.
Do you store the address I enter?
No. We only count how many checks run per day, not which pages. The address only sits in the link of your result, so you can share it.