On-Page SEO Checker

The On-Page SEO Checker audits one URL against the on-page basics that search engines and AI answer engines rely on: the title tag, meta description, H1, indexing, canonical, HTTPS, word count, image alt text, structured data, mobile viewport, language, Open Graph tags and internal links. It also checks your target keyword and which AI crawlers your robots.txt lets in. It is a free on-page SEO checker; the first time you use it, we ask for your email to show the results.

It is for founders, marketers and developers who want a quick, honest read on a single page before or after publishing. Every check comes with the reason and the fix.

Key takeaways

• Fix the red rows first: noindex, a missing title, no H1, missing alt attributes or a missing viewport tag.
• One H1 per page, a 30 to 60 character title and a 120 to 160 character meta description cover most quick wins.
• A self-referencing canonical stops UTM and tracking links from splitting your signals.
• Check the AI crawler table. Blocking search bots by accident can keep you out of AI answers.
• A perfect score is a baseline, not a ranking. Content and trust still decide competitive queries.

How the on-page SEO checker works

When you click Check this page, our server fetches the URL once. The fetch is guarded: it only visits public addresses on standard web ports, follows at most 3 redirects, gives up after 5 seconds and reads at most 2 MB. It also requests your site’s robots.txt and llms.txt. We don’t store the page. Results for the same URL are cached for 10 minutes, and each visitor can run 10 checks a minute and 100 a day.

The server extracts plain facts from the HTML. Your browser then turns those facts into checks.

Check	Good	Improve	Fix
Title tag	30 to 60 characters	Shorter or longer	No title
Meta description	120 to 160 characters	Shorter or longer	None
H1 headings	Exactly one	Two or more	None
Indexing	No noindex in meta robots or X-Robots-Tag		noindex found
Canonical	Points to this URL	Missing (one that points elsewhere shows as a Note)
HTTPS	Final URL is https		http
Word count	300 or more	Under 300
Image alt attributes	Every img has an alt attribute		Any img without one
Structured data	JSON-LD or microdata found	None found
Mobile viewport	Viewport meta tag present		Missing
Language	lang attribute on html	Missing
Open Graph	og:title and og:image both set	Either missing
Internal links	3 or more	Fewer than 3

Two details are worth knowing. An image with an empty alt attribute (alt="") passes, because that is the correct way to mark a decorative image; only images with no alt attribute at all fail. And the word count includes your menu and footer text, so a thin article inside a big template can still pass.

Target keyword. If you enter one, the tool looks for that exact phrase, ignoring case, in the title, the H1, the meta description and the first 100 words of visible text. A miss in the title, H1 or first 100 words is marked Fix; a miss in the meta description is marked Improve. On many sites the first 100 words include the menu, so a miss there may just mean your intro starts lower down.

AI search readiness. The tool reports whether robots.txt and llms.txt exist, and reads your robots.txt rules for nine AI crawlers: GPTBot, OAI-SearchBot, ChatGPT-User, ClaudeBot, Claude-SearchBot, PerplexityBot, Google-Extended, Applebot-Extended and CCBot. Each shows as Allowed, Blocked or Partly blocked. These rows are Notes and don’t change your percentage.

You also get your heading outline, the first 30 headings indented by level, which is the fastest way to spot a skipped level or an empty heading.

How to read your results

The headline number is the share of on-page checks marked Good. With 13 scored checks, each one is worth about 8 percentage points.

Checks passed	What it usually means	What to do
100%	The basics are in place	Spend your time on the content and on links
85% to 92%	One or two gaps, often Open Graph or a missing canonical	Quick fixes; most take minutes in your CMS or SEO plugin
70% to 84%	Several gaps, often title length, meta description and H1	Fix these before promoting the page
Under 70%	Template-level problems	Check your theme or CMS; the same issues are likely on every page

Look at the Fix rows before the percentage. One noindex makes every other check irrelevant, because the page can’t appear in search at all.

A worked example

Here is a typical service page we might check, with the target keyword “B2B SEO services”:

• Title: “B2B SEO Services for SaaS, Fintech and eCommerce Companies in 2026” is 66 characters, marked Improve.
• Meta description: none, marked Fix.
• H1: two, because the theme wraps the logo in an H1 as well as the page title, marked Improve.
• The other 10 on-page checks pass.

That is 10 Good out of 13 scored checks: 10 / 13 = 0.769, which rounds to 77%.

The fixes: cut the title to “B2B SEO Services for SaaS and Fintech” (37 characters), write a 150-character meta description, and change the logo from an H1 to a plain link. Now 13 out of 13 pass: 100%. Each fix moved the score by about 8 points, and none of them touched the content.

The keyword section would then show the phrase in the title and H1, and point out whether it also appears in the description and the opening paragraph.

What Google says about titles and descriptions

• 87% of the time, Google uses the HTML title element as the title link in results, up from around 80% before a 2021 change Google Search Central Blog, 2021

Google’s own documentation says there is no limit on how long a title element or a meta description can be, but that both are truncated in results as needed, typically to fit the device width. It asks for descriptive and concise titles, and says it uses your meta description for the snippet when it describes the page more accurately than text taken from the page. See Google’s guides to title links and snippets.

So the checker’s ranges are not Google rules. They are practical ranges that keep your key words visible.

Best practices

On-page basics for every page

• One H1 that says what the page is about, with the main keyword in natural words.
• A unique title with the keyword near the start and the brand at the end.
• A meta description written for the click: what the reader gets, in about 150 characters.
• An answer-first opening paragraph: two or three sentences that answer the query directly.
• Alt text that describes what each meaningful image shows; empty alt for decorative images. See our alt text guide.
• JSON-LD schema markup for the page type: Organization and WebPage, plus Service, Article or Product where it fits.
• A self-referencing canonical, so tracking parameters point back to the clean URL.
• At least three internal links to closely related pages, with descriptive anchor text.
• A deliberate robots.txt policy for AI crawlers.

Common mistakes

Myth vs fact

MythYou must keep titles under exactly 60 characters or Google will penalize you.

FactGoogle says there is no length limit. Long titles are simply cut off to fit the screen. Keep the important words first and avoid stuffing; the length range is about readability, not a penalty.

Myth vs fact

MythAdding schema markup will boost rankings.

FactStructured data helps search engines understand a page and can make it eligible for some rich results. It doesn't replace helpful content, and markup that doesn't match what's on the page can make it ineligible.

Myth vs fact

MythBlocking every AI bot in robots.txt protects your content with no downside.

FactSome of these crawlers fetch pages so AI search products can cite them, while others collect training data. Blocking all of them can keep you out of AI answers. Decide bot by bot.

Other mistakes we see often:

• Two H1s from the theme. Logos, sliders and widget titles are common culprits. Check the heading outline.
• A noindex left over from staging. It is the single most damaging line on a page, and it often survives a launch.
• The same title and description on every page. Usually a CMS default. Each page needs its own.
• Canonical pointing to the home page. Some plugins do this by mistake, telling Google every page is a duplicate.

A noindex left over from staging is the mistake I check for first after any launch, because it can quietly undo months of work. I'd rather spend two minutes running each key page through a checker than find the problem weeks later in Search Console. Make the check part of your launch list, not something you do when traffic drops.

What the checker can’t see

The checker reads one page’s HTML as our server receives it. It does not run JavaScript, so content that only appears after scripts run won’t be counted. It can’t judge whether your content is the best answer for the query, and it doesn’t see backlinks, brand mentions or field page speed. For speed, use PageSpeed Insights.

For crawl errors, redirect chains and site-wide issues, you need a full crawl. Our technical SEO service covers that, and our B2B SEO strategy guide shows where on-page work fits in a plan that leads to pipeline.