InsightsAEO/GEO

The 5 questions AI buyers ask before choosing a provider

Watch how people actually use assistants when they’re about to spend money and a pattern appears fast. The wording varies by category — lawyer, plumber, MSP, med spa — but the underlying questions don’t. There are five, they arrive in roughly this order, and each is decided by different evidence. This is the map of the conversation your business is either in or missing.

1. “Who are the best [providers] near me?”

The opening move and the highest stakes: the shortlist question. Engines answer it from corroboration — the businesses that trusted sources agree exist, serve the area, and are well regarded. Decided by citation presence, listing consistency, and review strength. If you win nothing else, win this one; it’s the question our leaderboard measures month after month.

2. “What should [this service] cost?”

Budget-setting. Engines synthesize from whoever publishes real numbers — which is why the providers with transparent pricing pages get quoted while “contact us for a quote” businesses watch the price anchor get set by competitors. You don’t need a full price list; ranges and typical-project examples you can stand behind are enough to be part of the answer.

3. “How do I choose between them?”

The criteria question — “what should I look for in a…”. Engines assemble checklists from expert content, and the providers whose articles supplied the criteria enjoy a quiet advantage: the buyer walks in pre-agreeing with their worldview. This is the strongest argument for publishing genuinely useful guidance rather than brochure copy.

4. “Is [specific business] any good?”

The verification question — buyers run it on referrals, on the shortlist from question one, and on you. The engine summarizes your reviews, recency and reputation; contradictions between sources surface here as hedged, lukewarm answers. The fix is boringly familiar: consistent facts, steady reviews that name service and city, and a public response to anything that deserves one.

5. “Who should I avoid?”

The question nobody optimizes for. Engines are cautious with it, but sustained negative patterns — review floods, unresolved complaints, coverage — do get reflected. You can’t suppress the answer; you can only be a business it has nothing to say about. Reputation hygiene is AEO too.

Winning the whole conversation

The five questions form a funnel, and evidence built for one compounds into the others: citations win the shortlist, published numbers win the budget anchor, expert content sets the criteria, review discipline survives verification. The playbook sequences the work; a free scan shows which of the five you currently win, question by question, engine by engine — because the honest starting point is knowing which parts of the conversation are happening without you.

Quick answers
Should I optimize for all five at once?

No — in order. The shortlist question routes the most buyers, and its evidence (citations, consistency, reviews) feeds every later question anyway.

Do buyers really phrase things this way?

The phrasings vary endlessly; the intents don’t. Our scans template real buyer phrasings per category — the same question set every month, so movement is comparable.

What if my category’s buyers don’t use AI yet?

Adoption curves only run one direction, and the evidence that wins AI answers — consistency, reviews, quotable content — wins referrals and search today. The work isn’t speculative even if the surface still is.

Letters to the editor
Write to the editor
The stories come from the count.

Every data piece is built on the live leaderboard scans — see who the AI engines actually name in your market.

Start the count.

request slip
TVX-______
your slip total
free snapshot — a human still reviews it
$0

a human reviews every request and every report before it sends — this queues your scan, it doesn’t fire a robot at your inbox · no newsletter unless you ask · protected by reCAPTCHA

You’ll get a confirmation now and the report within 24 hours.