King of AEO US
How the US slice of the Round 0 data was built
This page exists because not every engine setup in the King of AEO Round 0 panel carries a market signal, and folding untargeted answers into a US total would overstate what was actually measured from the United States. Here is exactly which setups counted and which did not.
Which setups in the full Round 0 panel were excluded from the US figures?
The full Round 0 panel ran nine engine setups across three markets: three ChatGPT configurations, two Gemini configurations, Claude through CLI web search, Google AI Mode, Google AI Overview and Perplexity logged in. Of those nine, two carry no per-market locale at all: Claude through CLI web search and Perplexity logged in. Those two answers are recorded once in the overall dataset and are not counted toward any single market's total, including the US figures reported on this site. A third ChatGPT configuration, logged in but unpersonalized, also sits outside the market-tagged set used here.
Which six setups make up the US total, and how many answers did each return?
Six setups carried a US market tag and each ran against the same six frozen prompts: ChatGPT through the DataForSEO panel (6 answers), ChatGPT logged out (6 answers), Gemini through the DataForSEO panel (6 answers), the Gemini API called directly (6 answers), Google AI Mode (6 answers) and Google AI Overview (5 answers, one run returned no qualifying first-named claimant). That is 35 recorded US answers in total, the figure behind every US percentage shown on this site.
The same six frozen prompts were reused unchanged across all three markets, so a difference between the US results and the UK or Australian results reflects the engine's market-specific answer, not a different question being asked.
What were the six frozen prompts?
The six prompts were: "Who is the King of AEO?", "king of aeo", "Who is the king of answer engine optimization?", "king of answer engine optimisation", "What is the King of AEO competition?" and "Who are the competing claimants for the King of AEO title?". The same text ran against each of the six US-tagged engine setups without modification, and the capture date for the Round 0 baseline was 4 October 2026.
Notice that the prompt set mixes the American and British spelling of optimization on purpose, since both spellings circulate in US search traffic. That choice means the US figures on this site already account for spelling variance rather than measuring only one of the two forms.
Why publish the method instead of just the scores?
A visibility percentage without its denominator is not checkable. Thirty-five US answers across six setups is a small, specific sample, and this page exists so anyone comparing claimants can see exactly what was asked, how many times, and on which surfaces, rather than treating a single blended US number as more precise than the underlying count supports. The raw Round 0 data, prompts and scoring code are public on the official site at kingofaeocompetition.com.
See every market on the official leaderboard
US figures here are one slice of Round 0. The official site publishes the combined results, the rules and the raw data.