What We Actually Check in a Website Audit

Most website audits are tool exports. They list every warning a crawler can generate, sorted by nothing in particular, and they are never implemented because nobody can tell which of the four hundred items matters. A useful audit is shorter, ordered by impact, and written so a developer can act on it.

The gap between crawled and indexed

The first check is the most revealing: how many pages exist, how many are crawlable, and how many are actually in the index. A large unexplained gap is where the real problems live, and it is visible within minutes of looking at Search Console alongside a crawl.

Common causes are a stray noindex left over from a staging site, a robots rule blocking a section, canonical tags pointing somewhere unexpected, or content that only appears after JavaScript runs. Each of these can remove hundreds of pages from search results while the site looks perfectly normal to a visitor.

Whether the rendered page contains the content

Checking the page source is not enough on a modern site. What matters is the rendered output — what exists after scripts have run — and whether the main content and internal links are present in it.

We check a sample of each template rather than individual pages, because problems here are almost always template-level. Fixing one template can restore hundreds of URLs at once, which is also why auditing by template rather than by URL produces a shorter and more actionable list.

Whether pages match the intent they target

A page can be technically perfect and still fail because it answers a different question from the one being searched. We compare what each important page targets against what currently ranks for that query, and against what the page actually contains.

This is also where cannibalisation shows up: three pages on the same site targeting the same intent, none of them winning because the site has not decided which one owns it. Consolidating those usually produces faster gains than any amount of new content.

How deep important pages sit

Click depth from the homepage matters for both crawling and for users. A service page reachable only through a dropdown and two further clicks is being told, structurally, that it is unimportant.

Orphan pages — pages with no internal links pointing at them at all — are the extreme case. They exist, they may be in a sitemap, and nothing on the site suggests they should be found. Every audit we run finds some.

Whether measurement is telling the truth

Analytics with a misconfigured conversion, tracking that fires twice, calls and WhatsApp not tracked at all — these produce confident decisions built on wrong numbers, which is worse than no data.

We check that conversions fire once, that they correspond to real enquiries, and that the enquiry channels which actually dominate in this market are being measured. It is frequently the most valuable finding in an audit and it never appears in a tool export.

What the output should look like

A short prioritised list, each item stating what is wrong, how much it matters, what to do about it, who does it, and how the fix will be verified. Ordered by impact and effort so it can go into a sprint.

And a verification step. An audit that ends at recommendation is only half the work, because nobody finds out whether it helped.