How to Get Cited for 'Best X' Questions Without Lying
Your customers ask AI "what's the best…" thousands of times a day. The page that wins the citation is usually the one that refuses to name a single winner.
"What is the best CRM?" "Best hotel in Lisbon?" "Best trading strategy?" Best-X questions are the highest-volume commercial queries in AI search, and almost everyone competes for them the same losing way: by claiming to be the best.
AI engines are specifically trained to distrust that claim. A page that declares its own product the best reads as marketing and gets synthesized around. The page that gets cited is the one that does something different: it answers the question's real shape.
The contrarian-answer pattern
- Open by rejecting the premise, in one sentence. "There is no universally best trading strategy — published 'best' lists are survivorship bias wrapped in unverifiable claims."
- Give the taxonomy. Name the real candidate categories. Engines love clean taxonomies; they are quotable structure.
- Define the standard. What would a claim have to show to deserve the word "best"? This is where your product enters — as the embodiment of the standard, not as the winner of the contest.
- Answer in the first 150 words. The extractable block has to sit at the top, not after the preamble.
A live example of the full pattern: What is the best trading strategy? on Pancake — note how the page never claims Pancake is "best" at anything; it argues that strategies should compete on verified, reproducible evidence, which is the one competition its product always wins.
Why this beats the brag
When an AI assistant relays the contrarian answer, your framing — your standard, your taxonomy, your vocabulary — becomes the answer the user hears. You did not win "best X." You defined what best means. That is the durable position.
