What is on-page SEO? Definition and the elements that matter

The usual split has three parts. On-page is what lives on the page. Off-page is what the rest of the web says about it, mostly through backlinks. Technical SEO is the layer underneath both: whether the page can be crawled, rendered, and indexed at all. The boundaries blur — a slow template is technical and on-page at once — but the split is useful because the three are fixed by different people.

What counts as on-page

Element	Its job
Main content	Answers the query. Everything else is packaging.
<title> element	The main input for the clickable headline in the results.
Headings	Structure for readers; context for engines.
Meta description	A pitch for the click, sometimes used as the snippet.
URL	Readable path, shown as breadcrumbs in results.
Internal links	Route authority and context between your own pages.
Anchor text	Tells readers and engines what the linked page contains.
Images and alt text	Discovery through image search; accessibility.
Structured data	Machine-readable facts about the page's entity.

Moz's list of on-page ranking factors puts content at the top for the same reason we do: it is the part users came for, and the rest of the list only helps a search engine understand content that is already worth understanding.

Content is the factor that decides the others

Moz frames good content as needing two properties: it must supply a demand, and it must be linkable. Both are worth taking literally. Supplying a demand means somebody is actually searching for this — a page nobody looks for cannot be rescued by markup. Being linkable means there is a reason to cite it: data, a tool, a real case, an opinion with a name attached.

Which demand a page answers, and in what format, is a question of search intent. Getting that wrong is the single most expensive on-page mistake, because no amount of heading tidying moves a product page onto a results page full of guides. How the copy is then written is covered under SEO copywriting.

The elements that appear in the search result

Two of them leave the page and show up in the results, which is why they get disproportionate attention.

The title link. Google generates it automatically from several sources: the <title> element, the main visual title on the page, heading elements such as <h1>, og:title content, and other large, prominent text. Its guidance is to give every page a specified, descriptive, concise title, avoid vague labels like "Home", avoid boilerplate titles that differ by one variable, and make it clear which text on the page is the main title. There is no character limit — the title link is simply truncated to fit the device width.

The snippet. Google says it primarily uses the content on the page to determine the snippet, and may use the meta description when that describes the page better than other parts of the content. So the description is worth writing well and is not worth agonising over: it is a click pitch, not a ranking factor, and a vague page body gives the engine nothing better to quote.

URLs and internal links

Google's starter guide is relaxed about URLs but specific about why they matter: parts of the URL appear in results as breadcrumbs, so words that help a human decide belong in the path, and random identifiers do not help anyone. Grouping topically similar pages into directories also lets Google learn how often each section changes and crawl accordingly.

Internal links are the on-page factor most sites underuse. They are how a new page gets discovered, how context flows between related pages, and how you tell an engine which of your own pages is the important one on a topic. Anchor text carries that meaning — Google calls it the text that tells users and Google something about the page you are linking to — so "our pricing" is worth more than "click here", and both are worth more than a link nobody put there.

What Google says you can stop worrying about

The starter guide has an unusually blunt section on this, and it saves a lot of wasted effort:

• Meta keywords — Google Search does not use the keywords meta tag.
• Content length — length alone does not matter for ranking; there is no magical word count, minimum or maximum.
• Keywords in the domain or URL path — hardly any effect beyond appearing in breadcrumbs, and the TLD only matters when targeting a specific country.
• Heading order — semantic order is excellent for screen readers, but Google says it does not matter for Search if headings run out of order, and there is no ideal number of them.
• Subdomain versus subdirectory — do what makes sense for the business.
• E-E-A-T as a ranking factor — the guide states plainly that it is not one.

None of these are permission to be sloppy. Heading order still matters for the humans using a screen reader; that is reason enough. They are permission to stop optimising things that do not move.

How Xerx does this

On-page work sits across two of our services, and we would rather say that than pretend one page covers it. The structural and markup half — titles, headings, internal linking, structured data, canonicals, and the crawl and render behaviour underneath — belongs to technical SEO. The writing half belongs to SEO content. Most audits we run find problems in both, and the fix list is ranked by impact rather than by which service it falls under.

We implement in the codebase. On a site we build, titles and descriptions come from the content layer rather than from a plugin field, headings follow the document outline, and internal links between related pages are part of the template rather than an afterthought — this Learn hub is a working example, with each term linking to its siblings and to the service it belongs to.

The diagnosis is free: the SEO report covers what is technically and structurally holding a site back, with every finding ranked and the fix named. Ongoing work runs inside SEO Care from €790 per month, in real logged hours, cancel monthly.

FAQ