On-page and technical checks
The on-page checks fetch your most visible pages and compare them with the queries people use. This page lists what is checked, how items are rated and, importantly, which thresholds come from Google and which are the tool’s own conventions.
✓ Checked against Google documentation and the workspace code · 7 October 2026
What the on-page checks do
The checks fetch your pages and compare what they contain with what Search Console says people search for. On-Page SEO compares each page’s leading queries with its title, meta description and H1. Full On-Page Check adds a technical audit of the same pages. Link Crawl follows your internal links from a start URL. The result is a prioritized list of things to review, not a score.
Full On-Page Check
- Choose the minimum impressions (50, 100, 500 or 1,000 over 28 days). The tool takes up to 1,000 pages ranked by impressions.
- Press “Run check”. Each page is requested once. The page records the HTTP status, canonical, meta robots and X-Robots-Tag, word count, internal and external links, structured data types, images without alt text, the viewport tag and whether the URL is in the XML sitemap.
- Read the list. Filter by severity, category (Canonical, Content, Headings, Images, Keywords, Links, Meta, Sitemap, Technical), HTTP status, minimum words and sitemap membership. Export as CSV with one row per issue or one row per query.
If your site blocks the crawler, allow the user agent SEOKitBot/1.0. The crawler respects robots.txt, and a page it is not allowed to fetch is reported as blocked.
The checks and how they are rated
| Check | Severity |
|---|---|
| HTTP error status | critical |
| Missing title, missing H1, redirect status, orphan page, a top query missing from the title | high |
| Missing meta description, missing or different canonical, missing viewport, blocked by robots.txt, non-indexable page in the sitemap, broken internal link, a top query missing from the H1 | medium |
| Title over 60 characters, description over 160, several H1s, under 300 words, images without alt, indexable page missing from the sitemap, a top query missing from the description | low |
The severities are the tool’s own judgement of how much attention an item deserves.
What is a Google rule and what is a convention
Several thresholds are conventions, not Google limits. Google says there is no limit on how long a title can be and that the title link is truncated to fit the device width (title links). It says the same of the meta description, and that snippets are mostly built from the page content itself (snippets). Google’s starter guide also sets no ideal number of headings. So “title over 60”, “description over 160”, “several H1s” and “under 300 words” are prompts to look, rated low for that reason, and not defects to fix by rule. Fix them when they hurt readers or make the result hard to understand.
Query matching
For each page the tool takes the queries with at least 100 impressions, up to 10, and checks whether each appears in the title, the description and the H1 as a whole phrase, ignoring case and punctuation. A miss is listed as “keyword potential”. Treat it as a question, not an order: a page can serve a query well without repeating the exact words, and Google rewrites titles when they do not describe the page (see H1 tag).
Link Crawl
- Start URL: the property root, or another URL on the same domain.
- Depth: 1, 2 or 3 link hops from the start URL.
- Pages: 50, 150, 500 or 1,000.
- Background: large crawls continue after you close the page, and you can cancel one.
The crawl reports the same categories plus indexability, international and structured data. An orphan page is one with no incoming links among the pages crawled, so a shallow crawl can call a page an orphan that is linked from deeper in the site.
Reading the results honestly
- The crawler’s HTTP answer is not Google’s. A firewall’s 403 to the bot does not prove the page cannot be indexed, and the tool does not count it as non-indexable. Confirm with URL Inspection.
- Raw HTML only. Content added by scripts after load is not seen.
- Start from critical and high, then look at a few low items rather than all.
If something looks wrong
| Symptom | Cause and fix |
|---|---|
| Many pages show HTTP errors | a firewall may block the bot; allow SEOKitBot/1.0 |
| Pages “blocked by robots.txt” | your robots.txt disallows them for this crawler |
| A page is called an orphan | no crawled page links to it; raise the depth or the page limit |
On-page checks, answered
Does Google require a 60-character title?
No. Google says there is no limit and truncates to fit the device width.
Should every page have exactly one H1?
Google sets no required number of headings; the tool rates several H1s low.
Does a pass mean the page will rank?
No. The checks find problems; they do not predict rankings.
Continue
Next, see title tag rewrites and the H1 tag.