Brand Signal

API and MCP for your brand data

Ask Brand Signal from Claude or ChatGPT, or bring survey data, conversations and AI answers into your own reporting, on the same competitors and demand spaces as the platform.

What the API and MCP are

Odyssey has both. The API gives data teams and agencies read-only access to Brand Signal, so survey data, conversations and AI answers can run in your own reporting and be joined with your own sales or media data.
The MCP server connects AI tools such as Claude and ChatGPT to Brand Signal. Your team asks in plain language, and the tool fetches the figures from your Signal, including the demand-space maps.
MixerMind is the assistant inside Brand Signal. The MCP lets your team ask the same data from the AI tools they already use.

What you can reach through the API and MCP

The API and the MCP reach the same data: the Signals, markets and competitors you see in Brand Signal, read only. The API feeds your data stack, and the MCP answers questions together with Claude or ChatGPT.

Awareness, consideration, preference and the other tracking measures, by market and period.

The demand spaces, their size and where each brand sits on the map, ready to draw.

Results by the dimensions in your survey, such as market, region, age and demand space.

Mentions, first places and cited sources for every brand in your competitor set, per engine and demand space.

Share of conversation, reach and sentiment for every brand in your competitor set, with paid and organic mentions kept apart.

Questions teams ask from their own tools

Ask in Claude or ChatGPT, or run the same queries from your own data stack.

  • 01

    Which demand space does Nova own in Sweden, and which competitors sit closest?

    Survey
  • 02

    For which needs is Nova named first in AI answers?

    AI answers
  • 03

    Do AI answers place Nova in the same demand space as buyers do?

    Survey and AI answers
  • 04

    How much of the category conversation is about Nova, with paid and organic mentions apart?

    Conversations
  • 05

    Draw Nova's positioning map for the UK with its five closest competitors.

    Survey
  • 06

    Join our monthly sales by region with consideration by region from Brand Signal.

    Your own data
  • 07

    Put Nova's first places in AI answers next to our media spend per market.

    Your own data

Examples with a fictional brand, Nova.

Security and data handling

Read-only access, with your Brand Signal login or your organisation's API key. Odyssey never uses your data to train AI models.

The MCP uses each user's own Brand Signal login and permissions. The API uses a key issued to your organisation.

The API and MCP read data. They cannot change it.

Every query is logged.

The API and MCP run on Google Cloud in the EU.

Never used by Odyssey to train AI models.

Through the MCP, answers go to the AI tool you connect, under your agreement with its provider.

Queries on your Brand Signal data.

Questions about the API and MCP

Yes. Brand Signal has an API that gives your data team read-only access to survey data, conversations and AI answers for the Signals you have access to. Talk to us to set it up.
Yes. The Odyssey MCP server lets AI tools such as Claude and ChatGPT query your Brand Signal in plain language. You sign in with your Odyssey login, and the tool can only read what your account can read.
Claude and ChatGPT are supported today. Generally the MCP works with any AI tool that supports MCP, but enterprise-grade tools and models work best. Ask us about other tools you want to connect to our MCP.
The API and the MCP give access to the same data: survey data, including the brand funnel, demand spaces and positioning maps; conversations, including share of conversation, reach and sentiment; and AI answers, including mentions, first places and cited sources.
MixerMind is the assistant inside Brand Signal, where an independent model checks each figure before you see it. The MCP lets your team ask the same data from the AI tools they already use and combine it with other sources there.
It depends on what you want to do with the data, and most teams combine them. Brand Signal is the place for analysis and interpretation. The API brings the measures you follow into your own dashboard, next to your other data. The MCP lets your team ask quick questions in Claude or ChatGPT. Talk to us and we will help you find the setup that fits.
You get the data next to your own sales, media and other sources, shown the way your teams already work. In Brand Signal, the views are built for your study: low bases are flagged, and periods and rolling averages are calculated the same way in every view. Through the API you get the same figures, with the value, the base and a flag when the base is low, but your dashboard decides how they are shown, which periods are compared and whether anything is recalculated. Those rules then need to be built in on your side.
Every answer the MCP returns carries rules for how the figures should be read, for example not to calculate new measures or differences, and to say when a base is low. We cannot see what your AI tool writes, and it can ignore those rules. The check MixerMind runs on every answer in Brand Signal does not follow the data into your tool, so check key figures in Brand Signal before you share them.
No. The API delivers aggregated results, with the value, the base and a flag when the base is low. Respondent-level answers stay in Brand Signal, where you can cut the data in new ways.
Your IT team builds the integration, with documentation and help from us. Survey data is more complex than, for example, media spend, because every figure is tied to a question, an answer option, a market, a period and a base. Those parts need to be handled correctly for your dashboard to show the same figures as Brand Signal. New data is available through the API as soon as a new wave is published.
Odyssey never uses your data to train AI models. When you use the MCP, the answers are processed by the AI tool you connect, under your agreement with that provider.
Only in AI tools used under a business or enterprise account where the provider's use of your data for model training is excluded by contract or switched off. Our terms for the API and MCP require this. Odyssey never uses your data to train AI models.
The API documentation is available to clients after login.
API and MCP access are part of Brand Signal. Talk to us about setting it up for your team.

Bring Brand Signal into your own tools

Nova and the other brands are fictional. Figures are illustrative.

WhoSverigeWhatNovaWhenSep 2026 · Smoothed (3)ShowStandardSave this view
Dark mode

MixerMind

I help you analyse your brand signals and find market insights.

Ask MixerMind about Brand Signal…

Send

[email protected]
A House #501, Stadsgården 6
116 45, Stockholm, Sweden
Privacy Policy
© All rights reserved