Skip to main content

Overall score

Each scanned HTML page gets an Overall score from 0 to 100. It blends five pillars: Meta & Head, Content, Social, Schema and Accessibility. Each pillar is scored from 0 to 100 on its own, so you can see which part of a page needs work.

MarkupCheck is a technical audit: each pillar scores whether the page's markup meets the standard, whether or not a search engine uses that check for ranking.

  • Images are scored separately and are not part of Overall.
  • UCP (Commerce Protocol), web app manifest, and favicon are property-health signals (Found / Missing), not Overall inputs.
  • The average score of a scan is the mean Overall of all its HTML pages.

Meta & Head pillar

Checks what search engines and browsers read first: the HTTP status, the page title, the meta description and the canonical URL, how scripts in the head load and whether the page text is there without JavaScript, whether links to other sites still work, and whether the HTML fits in the part Googlebot reads.

Title and description width

Search results cut titles and descriptions by width, not by a number of characters: 60 narrow letters (i, l) fit where 40 wide ones (W, M) do not. MarkupCheck measures them in pixels, the way search results show them, and falls back to characters for scripts it cannot measure.

Security headers, links that open a new tab without rel="noopener", long or messy URLs, heavy images and redirects are listed as advice. They cost no points.

Content pillar

Checks the page's structure and text: an <h1>, a heading order without skipped levels, enough text on pages meant to be read (articles, guides and service pages), and alt text on content images. Contact, pricing and category pages are short by design, so they are not marked down for it.

Heading checks

Headings are checked in the main content, so theme widget titles in a footer or sidebar never break the order. The issue names the jump, for example h2 “Pricing” → h4 “Monthly” skips h3. Long text without subheadings, more than one h1 and empty headings are listed as best-practice advice.

Social pillar

Checks the Open Graph and X (Twitter) tags that build the preview card when someone shares the page, and the share image itself: whether it loads and is big enough. The Social tab shows how X, Facebook, LinkedIn, WhatsApp, Slack, Discord, Pinterest, Reddit and other apps read the tags.

Schema pillar

Checks the structured data on the page, read from JSON-LD and from microdata (used by many Shopify, WooCommerce and older WordPress themes). It looks at the site graph (Organization, WebSite and BreadcrumbList) and at a type that says what the page is, such as Article, Product, Service or Event. Schema.org subtypes count as their parent, so a Restaurant counts as an Organization.

Recommended schemas by website type

The Website type you tick for a website (Ecommerce, Blog, Local business and so on, several at once) adds the schemas its pages should have, for example Product and Offer on product pages or BlogPosting and an author on articles. They show as checklists in the Schema tab of each page's details, one group per ticked type.

Which pages are checked

Each type checks its own kind of page. MarkupCheck finds them by the URL, by the schema already on the page and by og:type. On every other page the group says the type is not expected there and nothing changes.

How it changes the score

On those pages the website type's schemas count toward the Schema score, and the Issues list names what is missing. Changing a website's types updates the scores of its earlier scans too.

Accessibility pillar

Ten checks of the HTML that screen readers, keyboard users and browsers rely on: page language, image alt text, link and button names, form labels, field ids and autocomplete, heading order, unique ids and frame titles. Each one names the WCAG 2.1 success criterion it covers.

WCAG 2.1 AA readiness and risk band

The Accessibility tab groups the checks by the four WCAG principles: Perceivable, Operable, Understandable and Robust. Next to the score, a risk band (Low, Medium or High) sums it up. Automated checks find only part of the WCAG failures, so this is a readiness check, not a conformance audit.

Accessibility statement

Each scan also looks for the site's accessibility statement and checks it for four parts: the standard and level it aims for, a conformance status, known limitations and a feedback contact. The statement check is site-wide and not part of any score.

Images score

Not included in Overall. Each content image is checked for alt text and for a width and height, which keep the layout from jumping while the page loads.

The Links column shows how many links were found on the page. Page details list each link, where it sits on the page and what its anchor text says.

Page column

Shows the page URL path without the domain, for example /hu/adatvedelem. Hover a cell for the full URL and the page title. On desktop, drag the right edge of the column header to widen or narrow it; double-click the edge to reset. Click a score badge in the row to open its page details.

Status column

HTTP status code returned when MarkupCheck crawled the page (for example 200, 301, 404).

Load time

Time taken to fetch the page during the crawl, in seconds. It is not part of any score.

Crawl Snapshot

The Crawl Snapshot button in a page's details opens the HTML stored for that page during the scan (not a live re-fetch).

Score colors

Scores show in five colors, from Excellent and Good through Fair and Weak to Poor, on badges and in filters.

Property health (not Overall)

  • UCP — commerce protocol presence on the property (shop-flagged properties only)
  • Web app manifest — found/parsed (or missing)
  • Favicon — found via link tags or default probe

Signed-in users find the full method, with every check and how much it counts, in the app: Page scoring in the app.

← Back to MarkupCheck Dashboard