Skip to the main content
PicksYou

How it works

About two minutes from pressing the button to the report. Nothing is scraped, and nothing of yours is touched.

What a check does

We ask what a customer would ask

5 questions in your language, about your town or about what you sell online: who is best, who to call first, a comparison on price and reviews, who people trust, and how to choose. Each one goes to the assistants through the interfaces they publish.

The questions

For a business people look for in a town

  1. 1Who are the best dentists in Bristol?
  2. 2I am looking for dentists in Bristol. Which ones should I call first, and why?
  3. 3Compare the top three dentists in Bristol on price and reviews.
  4. and the rest of the pack for your kind of business.

25 questions for a business in a town, 25 for a business that sells online. Yours is asked from one list, never from both.

5 questions in a free check. 10 every Monday on Starter, 25 on Growth.

We read every answer for your name

A matcher looks for your name, your aliases and your website in each answer. When it is not sure, a person checks before anything counts. An uncertain match is never called a hit, so the count can be behind the truth but never ahead of it.

One answer, read

Named

Bristol has a lot of well reviewed practices, so "best" depends on what you need. For general and family dentistry, the ones that come up most often are Queen Square Dental Practice (city centre, strong reviews for check ups and hygiene), Clifton Smile Dental (Clifton, popular for cosmetic work) and Redland Dental Practice (family run, praised for nervous patients).

Also named: Queen Square Dental Practice, Clifton Smile Dental, Redland Dental Practice. Kept word for word, forever.

We show you every receipt, word for word

Every answer with your name marked where it appears. The count of answers that named you, with its margin. The reasons you were not named more often, each with what we found, why it matters and what to do.

The count, with its error

4of 15 answers named you

Run the same check tomorrow and the 4 could honestly be 2 or 6.

The waiting screen

While it runs you watch the answers arrive, assistant by assistant.

  1. Asking

    Question 2 of 5 asked.

  2. Answers arriving

    • ChatGPT 4 of 5
    • Perplexity 3 of 5
    • Google AI Overviews no answers yet
  3. Finished

    Picked in 4 of 15 answers (plus or minus 21 points).

Three frames of that screen as an example, not a check we ran. Your own counts are whatever the answers say.

Who is asked, and how many times

The free check and Starter ask 3 assistants. Growth asks 5 assistants: Grok and Claude join, once each per question.

Repeats are the measurement. One answer is one draw, so asking the same assistant the same question three times is what turns a lucky mention into a count with a usual range. The assistants that join at one pass are counted at one pass, and no sentence here pretends otherwise.

What comes back: a free check is 15 answers, a Starter Monday is 90 answers, a Growth Monday is 275 answers. Every one of them kept word for word.

Passes per question, by plan
Assistant Free check Starter Growth
ChatGPT 1 3 3
Perplexity 1 3 3
Google AI Overviews 1 3 3
Grok not asked not asked 1
Claude not asked not asked 1
Answers that come back 15 90 275

Every Monday

A plan turns the check into a series: the same questions, every Monday from 06:00 UTC. That is what makes the count mean something when it moves, because the only thing that changed between two Mondays is the world.

The mail carries the week's count and its usual range; the dashboard keeps every week readable, word for word, including after you cancel.

What you can conclude from one day

One sample on one day, with its usual range. Full method: How we measure.

What one day tells you

Whether the assistants know you exist today, what they say when they name you, and who they name instead. The count arrives with its usual range, because 15 answers is a sample and not a census: four of them today could be two or six next week by luck alone.

What one day cannot tell you

Whether a fix worked. A change in what the assistants say shows up in the weekly count over several Mondays, not in the next single check. That is what the weekly check on a plan is for: the same questions, every Monday, so a real change can separate itself from luck.

What we never do

No scraping

Answers come through each assistant's own published interface, the clean route. We never read their consumer apps, and an assistant with no such route is skipped rather than faked.

Nothing published without you

No change reaches your site or your profiles unless you approve that exact change, one click, logged. We suggest; you decide.

No position claims

A count of answers with its stated uncertainty, never a position number and never a promised recommendation. If a number here surprises you, How we measure shows how it was made.

No lock on your data

Cancelling the plan is two clicks from your billing page; one click stops the mail. Everything collected stays readable either way, and it all lives in the EU. The about page names the company and where it runs.

See what AI says about you

Check if AI picks you

Free. 5 questions, 3 assistants, 15 answers, about two minutes. No account, no card.