Being known is not being recommended: a five-time CEO's AI search test exposes the measurement gap
Tony Uphoff asked ChatGPT, Gemini, Perplexity and Google AI Overviews the questions his prospective clients ask, then counted how often the engines named the two businesses he launched in February. Across 59 answers, his newsletter Uphoff on Media appeared in none of 12 category answers, and his firm Uphoff Advisory appeared in one of 43. Every engine that answered "Who is Tony Uphoff?" knew his career.
Uphoff's conclusion is that models know who he is but almost never recommend what he has built. The engines also described him as CEO of Pipeline360, a role he no longer holds, because their sources were org-chart aggregators, executive databases and old press releases. He argues answer engines default to what the web has said most often, and most often means longest ago.
The test points to a split between being discoverable and being recommended. A brand that asks a model about itself by name gets an accurate answer and assumes it is covered, while a buyer's real question returns a shortlist it is not on. Uphoff plans to rerun the same prompts in January and publish the numbers either way.
He also found that when he asked what Demand-as-a-Service is, all three engines defined it correctly and cited Pipeline360's own pages without naming him, even though he helped create the strategy. The models credit whoever publishes the clearest explanation, not whoever had the idea first. A Corporate Ink study reported by eMarketer in May found only 29% of marketers actively track and measure AI visibility.
Reported by Search Engine Journal ↗
Topics: ChatGPT, Gemini, Measurement