Quick answer
Ranking tells you whether a page is retrievable for a query; being recommended tells you whether a model chose to name your brand in its answer. The two correlate but are not the same measurement, and a brand can be strong on one and weak on the other.
For twenty years, "are we visible" mostly meant one question: where do we rank for the keywords that matter. That question still matters, but it is no longer the whole answer.
What ranking actually measures
A ranking tells you that a page was retrieved and judged relevant enough to place among the results for a query. It says nothing about whether a reader clicked, trusted, or acted on it, only that the page showed up.
What being recommended measures
When an AI assistant answers "what's a good [category]" by naming your brand, it has done something a ranking never promises: it chose you, specifically, out of everything it could have said, and presented that choice as if it were a confident recommendation.
Why they can diverge
- A page can rank for a keyword without the brand being the kind of entity a model is comfortable recommending (no clear positioning, thin independent coverage)
- A brand can be strongly recommended by a model even where it doesn't rank particularly well for the matching keyword, if third-party sources describe it clearly and consistently
- Ranking is judged per page; a recommendation is judged about a brand as a whole, pulling from many sources at once
Neither measure replaces the other
What to actually do about it
- Keep tracking rankings; they still drive a large share of traffic and still feed what AI platforms draw on
- Add a second, separate measurement: a fixed prompt set, run on a schedule, checked for whether your brand is named at all
- When the two disagree (ranks well, rarely recommended, or vice versa), treat that gap itself as the thing worth investigating
Key takeaways
- Ranking measures retrievability; recommendation measures whether a model actively chose to name you
- The two are related but separate, and can move independently of each other
- A visibility program now needs both measurements, not just the one it already had
Frequently asked questions
If we rank number one, shouldn't we also get recommended?
Not necessarily. Ranking is judged per page and per query; a recommendation draws on everything a model has absorbed about a brand across many sources, so the two can disagree.
Cite this page
BrandWater AI Research. (2026, 16 October 2026). Ranking and being recommended are not the same thing. https://brandwaterai.in/blog/ranking-vs-being-recommended
How we work
Figures are dated and linked to their sources. Where none exist we say so. Read our methodology and AI transparency pages.