Nabil Asofi

Journal / ai news

Cited, not ranked: the split that broke old reporting

A page can rank fifth and get cited, or rank first and be skipped. Once you accept that, most visibility reporting stops making sense and something better replaces it.

The observation

In traditional search, position and visibility were near-synonymous. Position one meant the most clicks. Everything downstream — reporting, forecasting, agency pricing — was built on that equivalence.

AI answers broke it. Sources are selected for how well a claim can be lifted, whether the publisher carries authority on the topic, and whether the wider web agrees. Position is a weak input. Pages ranked fifth get quoted. Pages ranked first get summarised without attribution.

Why this is uncomfortable

Because it invalidates a comfortable story. “We moved you from position eight to position three” was a deliverable that could be produced monthly and understood by everyone in the room. It also happened to correlate with outcomes often enough to survive scrutiny.

Now you have two numbers that move independently, and neither one alone tells you whether you are winning.

What replaces it

Ranking, still. It has not stopped mattering — AI Overviews correlate reasonably with traditional rankings, and Google’s own position is that core Search quality is the foundation. Keep tracking it. Stop treating it as the whole picture.

Citation rate. Whether answer engines name you, measured as a rate across repeated runs because the answers are non-deterministic.

Mention framing. How you are described. Being named in a list of three tools is a different outcome from being named as the category leader. The citation rate does not capture this, and it is often the thing that actually moves deals.

Third-party presence. Where else you appear. A substantial share of citations come from sources you do not own — review platforms, communities, industry publications. If your entire visibility programme is your own site, you are optimising a minority of the surface.

The uncomfortable part

Citation and recommendation are different things, and citation is easier. Content can be useful enough to consult without being compelling enough to shortlist. In controlled studies, self-promotional comparison content has earned citations inside answers that recommended competitors — the publisher bought visibility for the competition.

Which means the work has moved. It is less “rank this page” and more “be the thing people recommend when the question comes up,” and that is influenced by what other people say about you at least as much as by what you publish.

What I do with this

Measure the four things. Separate citations from rankings in reporting, because merging them into a single visibility score hides the exact divergence you need to see. And spend real effort on third-party presence, because that is where the consensus is formed — and consensus is what a recommendation is made of.

Also on this subject