An answer engine is a system that responds to a query with a direct, synthesized answer instead of a list of links. Perplexity, ChatGPT with browsing, and Google's AI Overviews are the canonical examples.
Search engines rank documents and let the user synthesize; answer engines synthesize first and cite documents second. The shift compresses the funnel: the consideration set that used to be ten blue links is now three named brands in a paragraph.
For brands, the strategic difference is winner-take-most. Rank #4 on a search page still gets clicks; brand #4 in an AI answer frequently does not exist. The answer only named three.
Why it matters now
The answer engine compresses the funnel. The consideration set that used to be ten blue links is now three named brands in a paragraph. Search engines ranked documents and let the user synthesize; answer engines synthesize first and cite documents second, if at all. For brands, that changes what winning looks like.
It matters now because the format is winner-take-most. On a search page, rank four still earns clicks. In an answer, brand four often does not exist, because the answer named three and stopped. Presence in the answer is no longer a nice-to-have; it is the entire shelf.
How WhereDoIRank measures it
We treat the major answer engines as instruments to be read on a schedule. Each week we put the category's buying questions to ChatGPT, Claude, Gemini, and Perplexity, record the synthesized answers, and score which brands they name, how early, and how favorably, on a 0-100 scale per engine.
Because an answer engine returns a short list rather than a long one, the measurement is unforgiving in a useful way: a brand is either in the answer or it is not, and the close records exactly which, engine by engine.
A worked example
Imagine a buyer asks Perplexity, "What should a five-person startup use for CRM?" The engine returns two named products with one-line justifications and a few citations. A brand that would have placed sixth on a Google results page simply is not in this answer. There is no sixth slot.
For that brand, the answer-engine reality reframes the goal. Ranking somewhere on page one is not enough; the target is one of the two or three names the engine actually synthesizes into its reply.
Common misconceptions
An answer engine is not just a chatbot; Google's AI Overviews and citation-first tools like Perplexity are answer engines too, and the category is defined by the behavior, answer first, links second, not by the interface. It is also not a replacement for having good sources: the engine still retrieves and cites documents, so classic web presence feeds it rather than becoming irrelevant. And an answer engine is not neutral about position: unlike a results page with room to spare, its short list makes being named at all the decisive event.
An answer engine's short list is also not fixed. Ask the same question a week later and the named brands can change, because the engine synthesizes fresh from whatever it retrieves that day. That volatility is why a single answer is only a data point, and the close, not any one reply, is the record worth tracking.
How it connects
The answer engine is the environment AEO optimizes for and the driver of zero-click search, where the query resolves without a visit to any site. Google's AI Overviews put an answer engine on top of the largest search engine, which is why the SEO and GEO lines blur there. The brands an answer engine names are the raw material of AI visibility.