Topical authority is consistency

Topical authority is often sold as a score you can top up with more articles. I think of it as consistency. A site that covers one subject, keeps covering it, and keeps it in the same place gives search engines a simple expectation to confirm. Each new page on that subject lands where they expect it, linked from pages they already trust. Stay in your lane long enough and the lane becomes yours.

The same idea explains why migrations hurt. A move changes the addresses, the templates and often the subject mix all at once, and everything the search engines had come to expect has to be learned again. This post covers both halves. Deciding which pages a site should have in the first place is in the content and search intent guide.

Google's guidance on helpful content includes a list of questions for judging your own site. One asks whether the site has a primary purpose or focus. Two of the warning signs are producing lots of content on many different topics in the hope that some of it performs in search, and entering a niche topic area without real expertise mainly because you thought it would bring traffic.¹

A lane is not the same as a niche. Danny Sullivan of Google has said a blog can have multiple topics, and that the important part of that warning is the phrase about hoping some of it performs.² The test I use is the audience. A lane is the set of subjects the people you serve expect from you. A firm of accountants writing about tax, payroll, cash flow and company formation is in one lane. The same firm writing about celebrity net worth because it gets searched is out of it.

Google has described a topic authority system in one area. For news, it helps decide which expert sources are useful for queries in specialised areas such as health, politics and finance, using how notable a source is for a topic or location, its influence and original reporting, and its reputation. Google's advice to publishers was to provide great coverage about the areas and topics they know well.³ That system is scoped to news, but the advice reads the same for any site.

Outside news, Google's guide to its ranking systems says site-wide signals and classifiers are used and contribute to its understanding of pages. It also says good site-wide signals do not mean all content from a site will always rank highly, and that its systems are designed to work at the page level.⁴ A focused site helps each page. It does not carry a weak one.

The 2024 leak of Google's internal API documentation pointed the same way. Mike King's analysis described an attribute called siteFocusScore as capturing how much a site sticks to a single topic, and a site radius as capturing how far a page goes outside the core topic.⁵ He was clear that the documents do not show how features are weighted or whether everything listed is used,⁵ and Google cautioned against assumptions drawn from out-of-context, outdated or incomplete information.⁶ I read it as a sign that focus is measured, and nothing more.

Google's November 2024 update to its site reputation abuse policy is the clearest statement of what happens outside the lane. Google said it has systems to understand whether a section of a site is independent or starkly different from the main content, and treats such sections as if they were standalone sites. It warned site owners to expect traffic changes if sub-sections no longer benefit from site-wide signals.⁷

The spam policies give examples. One is a medical site hosting a casino page that is not integrated with the rest of the site. Another flags an article on a publisher's site because the content is unrelated to its editorial sections.⁸ Those are cases of renting out a site's reputation, and most businesses are nowhere near them. The ordinary version is a section added for traffic that the main site's audience never asked for. It is unlikely to borrow the authority the rest of the site has earned, and John Mueller of Google has pointed to a lot of totally unrelated content when asked about a site with indexing problems and ranking drops.⁹

Consistency covers where things live as well as what they say. Bing's webmaster guidelines ask for each URL to focus on a single topic, for URLs to stay stable over time with no unnecessary changes, and for consistent URL structures that consolidate signals.¹⁰ Bing also describes its judgement of quality as including the clarity of purpose of a site.¹¹

Crawling follows the same pattern. URLs that are more popular on the web are crawled more often, Google's systems recrawl pages often enough to pick up changes, and site-wide events such as site moves can raise crawl demand so the content can be reprocessed under its new URLs.¹² A search engine builds up a picture of a site from repeated visits: which sections exist, which templates hold which kind of content, which hub pages link to which articles.

That picture is what I mean by search engines rewarding what they find where they expect it. A new guide on your core subject, at a URL pattern the crawler already knows, linked from a hub that already ranks, is easy to place. A page on a new subject, in a new section, on a new template, has to be understood from scratch.

A migration breaks that picture all at once. Google's site move guide tells you to expect temporary fluctuation in rankings during the move, says it can take a few weeks or more for a medium-sized site's new URLs to start showing, and asks for redirects to stay in place for at least a year so all signals can transfer to the new URLs. It also advises changing only one thing at a time, with a domain move, a new CMS and a new layout done one after another.¹³

A migration guide on Bing's webmaster blog, written by Fili Wiese of SearchBrothers, makes the same point from the crawler's side: search engines need to recrawl a critical mass of known URLs before all authority has been transferred, and any migration will cause fluctuations and likely a temporary fall in visibility.¹⁴

Merges and splits are harder still, because they change what each site is about as well as where its pages are. John Mueller has said that site merges or splits generally take longer than one-to-one moves, with no fixed timescale, because reprocessing happens URL by URL.¹⁵ In my view the change of focus is part of why: the search engine is learning new addresses and a new answer to the question of what the site covers.

The recovery data is sobering, with a caveat. SALT.agency studied 1,052 domain migrations and measured how long monthly organic traffic took to return to a six-month baseline from before the move. The median was 304 days, and 13.9% had not recovered after three years.¹⁶ The study only picked up migrations where traffic fell by 60% or more, so it describes migrations that went badly.