picked

AI rank tracker: where AI answers put your brand, every day

Picked asks ChatGPT, Gemini, Perplexity, Claude and Google's AI answers your buyers' questions every day, and records whether each answer names you and in what place. Then it shows the brands named instead and the pages the AI read.

Picked is an AI visibility tool for agencies and B2B SaaS teams. It asks ChatGPT, Claude, Gemini, Perplexity and Google's AI answers the prompts your buyers ask, records which brands and pages they name, and writes the article built to become the source they quote.

  • 0 of 40

    answers named LaunchOnIt, for five prompts on four AI models

  • 33 of 40

    named Product Hunt, the brand ranked first for the same prompts

  • 32 of 69

    pages ChatGPT cited for a prompt on one day were cited again the next

LaunchOnIt, our own product, five buyer questions on ChatGPT, Gemini, Perplexity and Google AI Overviews, and ChatGPT's citations for ten questions, September 27 and 28, 2026, US.

What it measures

What does an AI rank tracker measure?

A Google rank tracker records one position per keyword. An AI answer has no list of ten blue links: it names a few brands in a paragraph or a list, and the next answer may name others. Picked records every answer and counts:

  1. Share of answers

    Out of all the answers to your prompts, the percentage that name you. It's the number to move and the one to put in a report.

  2. Average position

    Where you're named in the answers that name you: 1 means first in the list, 3 means two brands came before you.

  3. Rank among brands

    Where you stand against every brand the answers name, like "#4 of 31 brands named", with each tracked competitor's share next to yours.

  4. Sentiment

    How the answers present each brand, from 0 to 100: 0 warns against it, 50 is a neutral mention, 100 recommends it first.

  5. Sources

    The pages each answer cited, whose they are, and whether yours is one of them. That list says why a competitor ranks above you.

  6. Trend

    Each number by day, or as a 7-day average, with your articles and new listings marked on the line.

Not one position

Why does your AI rank change every day?

ChatGPT rewrites a question into one or more searches before it answers, and the pages it finds change. Google's AI Overviews and AI Mode fan a query out the same way. So the same prompt, asked twice, can name different brands in a different order.

That's why Picked asks each prompt again every day and reports a rate, not a screenshot. The 7-day average smooths the day-to-day swing, and the marks on the line show what moved it: an article going live, a page that started listing you.

A ChatGPT screenshot is an anecdote, not a metric.
There's no fixed position in an AI answer. The same prompt gets a different answer tomorrow: in our own runs, more than half the pages ChatGPT cited for a prompt changed from one day to the next. So an AI rank is a rate over many answers, asked every day, never one screenshot.

Share of voice

Is share of answers the same as share of voice?

Yes, for AI answers. Share of voice in advertising is your part of all the ads in a category; in AI search it's your part of the answers. Picked counts it per prompt, per AI model and per topic, and against each competitor you track. Some tools call it AI share of voice, share of model or visibility score. The arithmetic is the same: answers that name you, divided by all answers.

Real answers

What does an AI rank look like?

Five prompts LaunchOnIt, our own launch directory, wants to be named for, asked on ChatGPT, Gemini, Perplexity and Google AI Overviews: 40 answers, 76 brands named. Product Hunt ranks first, named in 33 of the 40. LaunchOnIt isn't named in any.

  1. Share of answers

    Product Hunt is named in 33 of the 40 answers, 83%. BetaList, second, in 25 of them, 63%.

  2. Across models

    All five brands at the top are named by each of the four AI models. Each model gave 10 of the answers, so a brand only one model names can reach 10 of the 40 at most.

  3. Your row

    LaunchOnIt is in none of the 40, so it has no position to average. The work starts with the pages those answers cite, and Picked lists them for each prompt.

Real answers, not a mock-up

LaunchOnIt's buyer questions5 prompts, 40 answers, 76 brands

Brands named

BrandAnswers naming it
  1. 1Product HuntNamed by all four models33
  2. 2BetaListNamed by all four models25
  3. 3Indie HackersNamed by all four models22
  4. 4Hacker NewsNamed by all four models20
  5. 5UneedNamed by all four models17
  6. ✓LaunchOnIt (You)Named (9.1%)

By AI model

Where do you rank on each AI model?

Every number splits by model, because each one reads the web its own way. In our runs, 61% of the brands named were named by one model only.

  • ChatGPT

    How Picked reads it
    The answer the real app gives, with web search
    How often
    Daily
  • Gemini

    How Picked reads it
    The answer the real app gives
    How often
    Daily
  • Google AI Overviews

    How Picked reads it
    The overview on Google's results page
    How often
    Daily
  • Google AI Mode

    How Picked reads it
    Google's AI Mode answer
    How often
    Daily
  • Perplexity

    How Picked reads it
    Through its API, with web search on
    How often
    Daily
  • Claude

    How Picked reads it
    Through its API, with web search on
    How often
    Less often than daily

Each plan includes some of them: 2 on Starter, 3 on Growth, 4 on Scale, chosen from the six. The pricing page has the rest of each plan. For one model in depth, read how ChatGPT picks what to recommend or how to track Google's AI Overviews.

After the number

What do you do when you rank low?

A rank tracker that stops at the rank hands the hard part back to you. In Picked, a prompt you lose leads somewhere, and the features page shows every screen on the way:

  1. 1

    See who took it

    Each prompt lists the brands named instead of you and the pages the answers cited.

  2. 2

    Get listed

    The roundups and review sites those answers cite, with who runs each one and a drafted ask. You send it.

  3. 3

    Write the page

    On Growth and Scale, Picked writes the article for the prompt: a brief you approve, a draft, a fact-check and a pull request on your site.

  4. 4

    Watch the rank

    Once the article is live, its page compares the answers before and after: how often they name you, and how many cite the new page.

When not to

When an AI rank tracker won't help

Honest answer: sometimes this isn't worth your time.

  • You want a guarantee

    Nobody can promise what a model will say tomorrow. Picked shows where you're missing and which pages the AI read instead.

  • You want one screenshot

    Ask ChatGPT yourself, it's free. A tracker is for asking every day and acting on the rate.

  • Your buyers don't ask AI

    A local trade where people buy by phone does better with classic local SEO.

Sources

How we checked

  • Our answers: LaunchOnIt, five buyer questions on ChatGPT, Gemini, Perplexity and Google AI Overviews, US, September 27 and 28, 2026. ChatGPT's citations, and the brands named by one model only: LaunchOnIt and Leado, five questions each, the same two days.
  • How the models search: OpenAI's help article on ChatGPT search and Google Search Central on AI features, read September 29, 2026.
  • What's built: the product as of October 12, 2026. Nothing on this page is planned.

Questions

Frequently asked questions

Short answers, each one checked against the product as built or the docs we link to.

A tool that asks AI assistants like ChatGPT, Gemini and Perplexity the questions your buyers ask, on a schedule, and records whether each answer names your brand, in what position, and which pages it cites.

A Google rank tracker records one position for a keyword on a results page. An AI answer names brands in text and changes from one asking to the next, so an AI rank tracker counts how often you're named across many answers and where you appear when you are.

ChatGPT, Claude, Gemini, Perplexity, Google AI Overviews and Google AI Mode. Each plan includes 2, 3 or 4 of them, chosen by you. Claude is asked less often, the others daily.

The percentage of AI answers to your prompts that name your brand, against your competitors. Picked calls it share of answers and shows it per prompt, per AI model and per competitor.

The free AI visibility check asks two questions about your category on ChatGPT and Google AI Overviews and shows who they name. It needs a work email, no account and no card.

Every prompt is asked again every day on each AI model in your plan, except Claude, which is asked less often. The dashboard shows each day and a 7-day average.

Find out what ChatGPT and Google AI say about you

Add your domain and the prompts your buyers ask. Unlimited seats for your whole team, and no promises we can't keep.