Semantic SEO Audit

The first check is whether the site agrees with itself about what it is. The source context is the business’s identity and purpose as a source; the central entity is the main subject the site should be known for. I read the homepage, about page, title tags, navigation and structured data side by side. If a moving company’s titles say “moving”, its blog is mostly home décor and its markup describes a generic local business, search engines receive three different stories, and every page below inherits the confusion.

Topical coverage is how much of a subject your indexed pages actually answer. I measure it by listing the attributes and sub-topics the central entity involves, then marking which ones the site covers well, covers thinly or misses. A storage operator might have a solid page for each facility and nothing on vehicle storage rules, business storage or what tenant insurance covers. Gaps like these explain why competitors with weaker sites still win: their coverage is more complete.

Cannibalisation is when two or more of your pages compete for the same search and neither ranks as well as one page would. I find it in Search Console’s query-level data rather than by comparing title tags: if two URLs swap positions for the same query over several weeks, search engines are unsure which one answers it. Dilution is the wider version of the problem, one topic spread thinly across many pages with none of them strong.

Entity signals tell search engines who is behind the site and whether its facts hold up. Knowledge-based trust, judging a source by whether its facts agree with what search engines already hold as true, starts with your own facts agreeing with each other. I compare your business name, address, phone number, founding details and key people across the site, your structured data, Google Business Profile and major directories. I also look for an entity home, the one page that defines the business authoritatively, and for sameAs links in your markup that connect it to your official profiles elsewhere. Fixes in this layer are carried out through entity SEO and schema markup.

Cost of retrieval is how expensive your site is for a search engine to crawl, render, understand and rank, and every wasted request raises it. I look for parameter and session URLs, redirect chains, internal links to broken or redirected pages, content that only appears after heavy JavaScript runs, oversized pages and slow server responses. For page experience I check Core Web Vitals, including INP (Interaction to Next Paint), using real-user data where it exists. Anything that needs development work is scoped as technical SEO tasks.

AI search readiness is whether your pages can be found, understood and quoted by AI answers. These systems break a question into many related searches, a process called query fan-out, then quote the passages that answer each part. That raises the importance of passage ranking, where one section of a page is ranked for a specific query on its own merits. I check whether your key answers exist as self-contained passages, whether headings say what each section answers, and whether robots.txt blocks AI search crawlers you would actually want. The ongoing work belongs to AI search optimisation.