Updated on 19 November 2026, by João Leite, Founder of murmur.

Branding in the AI era includes what AI models say about the brand when the buyer asks for a recommendation, and fame alone is not enough. The buyer asks ChatGPT, Claude or Google AIO, the AI summary at the top of the Google search page, which company to choose, and the answer names a few brands and describes each one. In the June 2026 Harvard Business Review, two marketing professors and a consultant who teaches at Georgetown report that, in their test with running shoes, the AI models often brought up Brooks, a relatively small brand, and Nike much less.

  • Branding in the AI era includes the AI answer. The brand also becomes what ChatGPT, Claude and Google AIO say about it when someone asks for a recommendation.
  • Fame weighs less in the recommendation, according to the HBR authors. In their test with running shoes, published in June 2026, Brooks appeared often and Nike much less.
  • The name weighs when the AI knows little. In the Peec AI study of August 2026, with English words, the meaning of the name entered the description only where the AI had little information about the company.
  • Recall and presence are two different calculations. Folha Top of Mind, by Datafolha, measures people’s recall; presence counts in how many AI answers the brand appears.
  • Positioning is written down. The sentence that says what the brand sells and to whom needs to be in text, on the site, and the same everywhere.

Branding in the AI era: the brand becomes what the answer says about it

In branding in the AI era, the brand also becomes what the answer says about it. Branding is the work of building what people think and say about a brand, and that work has gained a new reader: the AI the buyer consults before choosing. When someone asks which running shoe to buy, which system to subscribe to or which agency to hire, the answer chooses which brands to cite and with what words.

The brand’s description in the answer becomes part of branding. It says who the brand serves, what it does best and which category it belongs to. The tone of that description, whether the AI speaks well or badly of the brand, has its own guide, on how to know if the AI speaks well or badly of your brand.

In the HBR authors’ test, with running shoes, Brooks appeared more than Nike

In the test the authors report in the Harvard Business Review, with running shoes, Brooks appeared in the AI answers more often than Nike. The article, from 29 June 2026, is by John Gale, consultant and adjunct professor at Georgetown’s McDonough School of Business, Luca Cian, marketing professor at the University of Virginia’s Darden School of Business, and Luc Wathieu, marketing professor at Georgetown University. In their test, run on ChatGPT, Claude and Gemini, the authors write, in the Harvard Business Review article, that “the relatively small brand Brooks appeared reliably”, while Nike appeared far less consistently.

In the summary of the article, the authors write that “traditional brand-building advantages such as awareness, storytelling, and emotional appeal are becoming less decisive”, that is, that traditional advantages of brand building, such as recognition, storytelling and emotional appeal, are weighing less in the decision. The test covers one category, running shoes, and in it being the bigger brand did not put Nike in the answers as often as Brooks.

What weighs most in the answer is the brand’s presence on the internet, more than size and fame

What weighs most in the answer is the brand’s presence on the internet, how much is written about it, on its own site and outside it; size and fame, for the HBR authors, weigh less. The four reasons that make AI recommend some companies and leave others out are in the guide on the reasons behind AI recommendations.

AI links each brand to a category, and it is through that link that the brand enters the buyer’s questions that do not carry its name. The brand that AI links to what it sells appears when someone asks about the category, and how to read that link is in the guide on share of model.

The brand name enters the description when the AI knows little about the company

The meaning of the brand name enters the AI description when it knows little about the company. In August 2026, Peec AI, which sells AI monitoring, tested whether the meaning of one-word English names changes the description the AI makes of the brand, with a list of about 14 thousand English words evaluated by people. The conclusion of the Peec AI study on the brand name is that the effect of the name appears only where the model has little information about the company, and disappears when the brand is already known to it.

For branding, the study points to information, not to the name. In Peec’s reading, the way out for a brand that already exists is for the AI to gain information about the company, not a new name.

Top of Mind and presence in the AI answer are two different calculations

Top of Mind and presence in the AI answer measure different things: the first counts people’s recall, and the second counts AI answers. In Folha Top of Mind 2026, published on 29 August 2026, Folha de S.Paulo reports that Datafolha visited 171 municipalities to find the most remembered names in different categories. It is a survey with people, about the brand that comes to mind.

what is comparedFolha Top of Mindpresence in the AI answer
what it countsthe most remembered names in different categoriesin how many answers the brand appears, divided by the total number of answers
who answerspeople, in a Datafolha surveythe AI, when the buyer asks
how it is collectedin 2026, Datafolha visited 171 municipalitiesthe buyer’s questions put to AI models, repeated over time
what the calculation saysthe brand that comes to people’s mindswhat the AI delivers to whoever asks

Presence in AI is calculated another way. Presence is how many answers the brand appears in, divided by the total number of answers, a calculation the market also calls share of voice. One calculation does not replace the other: recall says what people keep about the brand, and presence says what the AI delivers to whoever asks.

What the brand team does in the AI era, in five steps

The brand team, in the AI era, makes sure the answer finds the brand described the right way. The steps:

  1. Write the positioning sentence with the buyer’s words: what the brand sells, to whom and where it operates.
  2. Publish the brand’s facts in text, on the site, with what it sells, to whom and since when, outside images and buttons.
  3. Use the same name and description everywhere, on the site, on LinkedIn and in directories.
  4. Ask the AI models the buyer’s questions and see how each answer describes the brand and the competitors.
  5. Track the same questions over time, to see whether the description changes after each step.

Step 3 belongs to the brand team in the GEO work, the work of making the company appear in the answers of ChatGPT and the other AI models, and the division between teams is in the guide on who owns GEO in the company. If the answer echoes what customers left on review sites, it is worth reading the guide on the weight of reviews.

murmur positions the brand in the answers: Fly Vet in front when ChatGPT recommends

murmur, a GEO agency in Brazil that places brands in the answers of ChatGPT, Claude, Google AIO and other AI agents, works where the brand is now decided in AI: in the answer the buyer reads when asking for a recommendation. It helps correct and position the brand the right way in AI models, with a person dedicated to the brand from the diagnostic onward, who studies the company and the sector, reads the results and follows each step closely.

Fly Vet, a veterinary marketing agency with proprietary technology and a murmur client, shows in its case study the brand positioned in the answer. In questions about hiring a veterinary marketing agency, ChatGPT put Fly Vet in first place in 30% of them, on 2 September 2026. None of those questions mentions Fly Vet by name: the buyer asked about the category, and the AI answered with the brand.

Frequently asked questions

Does a strong brand still matter when AI recommends?

It matters, but it carries weight differently: in the June 2026 Harvard Business Review article, the authors write that traditional advantages of brand building, such as recognition, narrative and emotional appeal, carry less weight. In their test with running shoes, Brooks, a relatively small brand, appeared often, and Nike much less. What enters the answer is what the AI finds written about the brand, on its site and outside it.

Does changing the name help a brand appear in AI?

In the Peec AI study from August 2026, with one-word English names, the effect of the name disappeared when the AI had information about the company. Peec writes that the way out for a brand that already exists is for the AI to have information about it, not a new name.

Do Top of Mind and AI presence measure the same thing?

No: Folha Top of Mind, by Datafolha, measures people’s recall, and in 2026 Datafolha visited 171 municipalities to find the most remembered names in different categories. Presence is a count of AI answers: how many answers the brand appears in, divided by the total number of answers. One calculation does not replace the other.

Who handles branding in AI, the brand team or the SEO team?

The brand team handles the company name and description, which need to be identical on the site, on LinkedIn and in directories, and the work brings other teams together. What each one delivers is in the guide about the division of GEO among teams.

Where to start

The first step is to see how AI models describe your brand today when someone asks for a recommendation. With your site’s URL, murmur’s diagnostic shows in less than two minutes, and without signing up, the AI answers to your market’s questions; the full report arrives by email. To see the brands AI models already recommend in each sector, murmur’s sector studies include the round for each market measured.

Read also

References