About SeeWAIS (See What AI Says)

Why this exists

People have started asking assistants the things they used to type into a search box. Best dentist near me. Who should I hire for this. Is this company any good. Put your own business in the blank. The answer that comes back names two or three businesses and silently omits everyone else. Unlike a page of search results, there is no second page.

Almost no small business has ever seen what those answers say about them, and it isn't for lack of caring. Checking properly means asking the same questions across several assistants, repeatedly, and reading every answer. That's an afternoon of work, and the result goes stale.

That's the gap. Not a strategy, not a retainer, not a dashboard. One look, done properly, at a price where finding out is not a decision you have to think about.

What you actually get

30 questions — the kind your customers actually type — put to 5 AI surfaces, 176 answers recorded and read. 3 of those questions get asked 3 times, because these systems are not deterministic and a single run of a drifting system is a coin flip. You get a PDF that is yours to keep, and a link that stays live for 12 months. Full detail is on how we measure, and a complete sample report is on the sample page. It is the whole report, not a screenshot of one.

What we won't do

We never publish your results

Your report lives on a private link and is served noindex. Your scores never appear in a public ranking, a leaderboard, a "top brands in AI search" index, or a marketing post. Several companies in this category run exactly those. We don't, and we won't — you paid us to tell you something, not to make you a data point.

We'll tell you when you're fine

One of the 4 verdicts this system can return is fine: nothing here is worth your money, don't buy anything, go back to work. It fires when the measurement says so. A product that can only ever discover that you need help is not a measurement. It's a sales call.

No subscription, and nothing to log into

$4.99, once. There is no plan, no seat, no renewal, and no dashboard demanding a weekly visit. You get the measurement, the evidence behind it, and a ranked list of fixes you can act on yourself.

We don't claim a trick that doesn't exist

There is no file you can upload and no markup you can add that makes an assistant recommend you — Google says so directly, and we quote them saying it on our FAQ. Anyone selling you one is selling you something else. What the report gives you is a measurement and the specific, boring things that actually moved it.

How to check our work

You should not take any of the above on faith, so none of it asks you to:

• The method is published in full — every engine, every question type, how answers get graded, and what the score does and doesn't mean. The version is stamped inside every report, so you can tell when it changes.
• The benchmark is published — the same pipeline run against 34 well-known brands. You know these companies. If our numbers about them look wrong to you, that tells you something about our numbers.
• A complete sample report is published, including the appendix where we write down what the method cannot see.
• Every figure on this website is read out of the same code that generates the reports. The site cannot claim we test something the pipeline doesn't test — it refuses to start if the two disagree.

Who runs it

See What AI Says is independently run and small. You'll notice there's no founder photo and no team page, and that's deliberate: we'd rather hand you the method than a headshot. A report is either right or it isn't, and everything you'd need to decide that is on the four links above.

What we do owe you is a way to reach a person, and terms you can read before you pay. Both are here:

• Email [email protected] — a person reads it. Quote your report ID if you have one.
• Terms, Privacy and Refund policy, all in plain language.
• Lost your link? Find my report gets it back without involving us.