This page is about one tool’s data: the Meta Tag Analyzer. For how these scores sit alongside those of the other two tools, and the distributions in full, see SEO Score Benchmarks — the methodology and privacy notes are there too.
Which tag is missing, and how often
The tool takes a URL, fetches the page’s <head> and runs ten checks: the presence and the length of the title and description, the canonical, the charset, the viewport, the language attribute, Open Graph and Twitter Card. It asks for no keyword — so what it measures is not how well a page matches a query but how properly it introduces itself. The percentage below is the share of runs in which the check did not pass.
| Check | Did not pass | Weight |
|---|---|---|
| Title length (30–60) | 52% | Partial |
| Meta description length (70–160) | 51% | Partial |
| Twitter Card | 39% | Partial |
| Canonical URL | 35% | Partial |
| Open Graph tags | 31% | Partial |
| Meta description present | 27% | Hard fail |
| Charset declared | 23% | Partial |
| HTML lang attribute | 11% | Partial |
| Mobile viewport | 9% | Hard fail |
| Title tag present | 4% | Hard fail |
The three items at the bottom of the table — the language attribute missing in 11% of runs, the viewport in 9% and the title tag itself in 4% — now read as a sign of a broken install rather than an oversight. A CMS ships all three by default; when they are absent the page was usually hand-written or the template was left half-finished.
The real problem is not absence — it is length
The two items at the very top of the list are length rules: the title is outside 30–60 characters in 52% of runs and the meta description outside 70–160 in 51%. The tag being absent altogether is far rarer — no title tag in 4% of runs and no meta description in 27%. That gap is the most useful thing in this dataset: nearly everyone writes a title, and about half write one that will not survive the search result. Fixing it takes no new content, only re-measuring the space you already have — which makes it one of the few SEO gaps that can be closed in an afternoon.
The category view lands in the same place: Title & description is the lowest-scoring category in 58% of runs. The average scan passes 7.2 of the ten checks and leaves 2.8 open items behind it, of which only 0.40 are scored as hard fails.
Sharing tags: the loss nobody notices
Open Graph tags are missing in 31% of runs, a Twitter Card in 39% and a canonical URL in 35% — and Social tags is the weakest category in 31% of runs. What these three share is that their absence shows no symptom on the site itself: the page looks normal and nothing moves in analytics. The difference only appears once the link is shared somewhere else, as an empty box with no headline and no image. A missing canonical is quieter still and costs more: wherever the same content is reachable from more than one address, you have left the choice of which one counts to the search engine.
Frequently asked questions
How long should a title tag be?
The tool targets 30–60 characters, which is roughly the width Google renders on desktop before it truncates. In this dataset 52% of runs fall outside that range — that counts both titles under 30 characters and titles past 60 that get cut off. A truncated title is not wrong, but it loses the end of its sentence at exactly the point where the click is decided, so putting the key phrase first is safer than leaving it last.
Is the meta description still necessary?
It is not a ranking factor, but if Google does not like yours it writes its own from the page — so not writing one just hands the job over. In this sample 27% of runs have no meta description at all, and once the ones that exist but sit outside 70–160 characters are added, 51% of runs miss the length rule. The scoring treats "absent" as a hard fail and "wrong length" as partial credit lost.
Do Open Graph and Twitter Card tags affect SEO?
Not rankings directly — they decide what the page looks like when someone shares it. They do come second on the weakness ranking, though: Social tags was the lowest-scoring category in 31% of runs. Open Graph tags are missing in 31% of runs and a Twitter Card in 39%. Without them, every link posted to a social network or pasted into a chat renders as a bare box with no headline and no image.
Why do the scores cluster at the extremes?
Because there are only ten checks, and a page’s tag layer tends to be either set up or not set up at all. 353 runs scored A+ while 303 scored F, with just 94 runs in the C band between them. Install a theme or an SEO plugin and most of these tags arrive together; install neither and none of them do.
The sample, the methodology, the anonymization and the CC BY 4.0 licence are all set out in the methodology section on the hub.