Why We Measure Authority, Not Just Rankings

Rank reports used to be simple to read: position three for that keyword, up from position seven. Everyone understood what that meant, and it felt like progress.

That number still matters. But on its own, it's started to mislead people more than it helps them.

Here's the problem. When an AI system pulls together an answer instead of showing a results page, being ranked #1 doesn't mean much if that page isn't the one the system actually trusts enough to draw from. A page can sit at the top of the results for a term while the AI-generated answer for that exact question cites a competitor three spots down. Same query, different winner, because the two systems are measuring different things — which is really the whole shift answer engine optimisation is about.

So we track something else alongside rankings. We call it authority, mostly because there isn't a cleaner word for it.

What authority actually means

Authority is a measure of whether other sources — sites, publications, and AI systems themselves — treat a page as worth pointing to, consistently, across more than one question. It's messier to define than a rank, which is probably why so few agencies bother measuring it properly.

A site can rank on page one for a keyword and still have almost no authority behind it — thin content that happens to match the query well, propped up by decent technical SEO. We see this a lot with newer sites chasing quick wins. And the reverse happens too: a page that isn't top of the results for any single term but keeps getting referenced anyway, because it's the source other people trust when they're building their own content.

Why rankings and authority behave so differently

Rankings are volatile because they're recalculated constantly against whatever else is currently competing for that query; authority is stable because it's built from a track record that doesn't reset every time a competitor publishes something new. An algorithm update, a stronger competing article, a seasonal shift in demand — any of that can knock a page down five places in a week, without the page itself changing at all. It's easy to panic over a drop like that, even when it has nothing to do with the content itself and everything to do with someone else's launch.

Authority takes longer to build and it's genuinely harder to fake, which is exactly why it's worth more. It also seems to be the kind of signal AI systems lean on when deciding what to cite, since they can't verify a live ranking the way a search engine can — they're relying on a broader read of whether a source has been reliable before.

How we actually measure it

When we assess a client's authority rather than just their rankings, we look at four things: whether other credible sites link to or cite the content; whether it's showing up as a source inside AI-generated answers, not just the traditional results page; whether the claims on the page would survive someone fact-checking them; and whether visibility is spread across a cluster of related questions rather than propped up by one lucky keyword. That last one tends to be the most honest signal — a page that only shows up for one exact phrase usually got there through optimisation, not trust.

We're not dropping rankings from the reporting. They're still real, still measurable, and still connected to traffic. But they're one input now, not the whole scoreboard. The thing we're actually trying to build for clients is simpler to say than to achieve: be the source that gets trusted, by people and by the systems increasingly answering on people's behalf — which in practice comes down to backlink quality and earned coverage.