Skip to content

Measured every day

Track if ChatGPT names you, and where

For every buying question you track, ChatGPT’s answer is saved every day with the order in which it names brands: whether it names you, where you sit in its list, which competitors come first and which sources it cites. Position is read from the answer itself, never estimated.

Free, no card needed: four questions to ChatGPT and Gemini. See the plans

Average position only counts the answers that name you inside a list, and says how many they are. A day without a measurement stays blank, never a zero.

How it works

How your ChatGPT position is read

  1. Questions your customers ask

    Write what your buyers ask before choosing, or accept the questions the platform suggests from your website. ChatGPT gets them as they are, in the web version a buyer uses.

  2. Position from the list itself

    When the answer has a list or a table, your position is the place of the first item whose heading names you. A narrative answer counts as a mention without a position: no ranking is forced.

  3. Day by day, with its sources

    Every answer is saved in full with its date and the sites it cites. You see when you move up or down and which sources back whoever is ahead.

Key features

Your position in ChatGPT's lists

Position is read from each answer's text, never estimated. This is what you see.

Position

Position from the list itself

When the answer has a list or a table, your position is the place of the first item whose heading names you. A narrative answer counts as a mention without a position: no ranking is forced.

Average position

It only counts when there is a list

Average position only counts the answers that name you inside a list, and says how many they are. A day without a measurement stays blank, never a zero.

Competitors

See who ranks ahead

The competitors ChatGPT names before you on each question, with their average position.

Sources

Check the sources that weigh

The pages it cites when it places you high or low, so you can decide where you need to be.

Check my brand for free
Platform

Everything else, in the same workspace

The same questions and answers feed every screen, so the numbers agree with each other.

  • Question tracking

    Your questions, measured every day, with their history.

    Source analysis

    The sites the answers cite, and your gaps.

  • Competitor analysis

    Who appears with you, and where they get ahead.

    Sentiment analysis

    How the answers describe you, sentence by sentence.

    Social listening

    The Reddit and Quora threads Google shows.

Console views with demonstration data for a fictional brand.

Coverage

Different answers. One place to understand them.

What each platform answers, with its date and sources, in one dashboard.

Grow measures 4 platforms every day: ChatGPT, Gemini, Google AI Overviews and Perplexity. Scale measures 6: ChatGPT, Gemini, Google AI Mode, Google AI Overviews, Microsoft Copilot (Spain only) and Perplexity. We measure their responses. Each question is measured from its brand’s country, in that country’s language: Spain, the United States, or Germany. Perplexity is measured through its API with web search, and the others in their web versions. Claude and Grok are measured every week for all your questions, on both plans and outside the daily AI platforms, through their APIs with web search: not their apps or a personalised account. Weekly balances cost and follow-up; their answers can vary between queries. Every answer keeps its date and sources so you can compare under the same conditions.

Check my brand for free
Before you start

Frequently asked questions

What people ask us most about ChatGPT.

Can't find your question? Write to us

  • What is a ChatGPT rank tracker?

    A tool that records how ChatGPT answers your buyers’ questions, with the place it gives your brand and each competitor, so you can follow the changes over time.

  • How is the position decided?

    Only when the answer has a clear list: a table with at least two rows or at least two consecutive items. Your position is the first item whose heading names you, and two brands in the same heading share the place.

  • Is the position the same for everyone?

    No. ChatGPT changes between sessions, days and users. That is why the same questions are measured every day and you see trends and averages, not one fixed rank.

  • Do you measure the ChatGPT app or its API?

    The web version a buyer uses, through a capture provider. Each question is measured from its brand’s country, in that country’s language: Spain, the United States, or Germany. We don’t use the API. If an answer arrives without web search or cut short, it is flagged.

  • What is saved from each answer?

    The exact question, the language and market, the date and time, whether web search was on, the answer as it arrives, the brands and their order, and the addresses it cites.

  • Which AI platforms does the platform measure?

    Grow measures 4 platforms every day: ChatGPT, Gemini, Google AI Overviews and Perplexity. Scale measures 6: ChatGPT, Gemini, Google AI Mode, Google AI Overviews, Microsoft Copilot (Spain only) and Perplexity. Each question is measured from its brand’s country, in that country’s language: Spain, the United States, or Germany. Perplexity is measured through its API with web search, and the others in their web versions. Claude and Grok are measured every week for all your questions, on both plans and outside the daily AI platforms, through their APIs with web search: not their apps or a personalised account. Weekly balances cost and follow-up; their answers can vary between queries.

  • What happens automatically?

    The plan measures your questions, finds conversations and prioritises next steps. Enable automatic preparation to receive drafts within your monthly allowance. You review and publish. It is paid by card, from $109 a month + applicable taxes, with no minimum term.

  • Do you guarantee I’ll appear in ChatGPT?

    No. Nobody controls what these systems answer. The platform measures what they answer, with dates and sources, so you can see what changes; it doesn’t promise mentions, positions, traffic or sales.

Your next step starts here

Measurements in ChatGPT, Gemini, Copilot, Perplexity and Google (plus Claude), evidence-backed priorities and included drafts to review. Choose your capacity and get started.

See plans

Start with a free check of your brand