53% of consumers already use AI to shop. Does your brand show up?

The comparison between brands and stores, which used to happen on search engines, websites and in conversations with salespeople, is moving elsewhere. McKinsey research published on October 28, 2026, shows that more than half of consumers already use artificial intelligence to shop. The consequence for sellers is that, for a large share of consumers, the shortlist starts taking shape in a conversation with a machine, before any contact with your team.

What happened

McKinsey surveyed about 21,000 consumers in 11 markets, Brazil among them, in June 2026. On average, 53% said they had used AI to shop in the previous three months. Among younger consumers, the figure rises: 62% of Gen Z and 63% of millennials.

Use goes beyond price research. Nearly a third of shoppers who use AI named these tools as the biggest influence on their most recent purchase. And 81% said AI does a good job of filtering and narrowing down options.

The purchase itself is also starting to go through the machine. In the United States, where there is a comparison with the previous measurement from December 2025, the share of AI users who completed a purchase through the tool itself rose from 10% to 17%, averaged across the categories measured. Across all countries, 14% of AI users have already done so.

Why it matters

The figure that weighs most is influence, ahead of completed purchases. If AI narrows ten options down to three, those left out of the three lose without knowing they were in the running, before any contest over price or negotiation.

This changes the nature of the funnel. For years, companies measured the top of the funnel by what they could see: visits, clicks, inquiries. Now part of the decision happens in a place that doesn't show up in the report, and the customer arrives with a preference already formed, or simply doesn't arrive.

There is also an effect on price. A tool that compares offers side by side rewards those who explain clearly what they deliver and penalizes generic proposals. When the difference isn't clear, the comparison tends to fall back on the easiest criterion to measure, which is price.

What changes for you

For the owner: the company's first impression now includes what AI says when someone asks about your market, alongside the website and referrals. It deserves to be treated as an asset, with an owner and ongoing monitoring, instead of as a marketing detail.

For sales leaders: the incoming lead may have already ruled out competitors or already have a favorite. The first conversation needs to start from what the customer already knows and correct what they got wrong, instead of starting from scratch. In early contacts, ask where the customer did their research before reaching out.

For operations leaders: inconsistent information across the website, catalog, price lists and materials turns into errors in the machine's answer. Outdated prices, an old portfolio and vague descriptions knock the company out of the comparison. Organizing that information base has become part of the sales process.

What changes in Brazil

Anyone selling to Brazilian consumers should already treat the AI answer as a new storefront. For some shoppers, the comparison starts with an assistant, and those who don't show up in that conversation never make it into the contest. Organizing your own information so these tools can read it costs little compared with the risk of being left off the shortlist.

The country is among the 11 markets McKinsey surveyed, although the study does not highlight the Brazilian figure. The local share may be higher or lower than the average. Either way, the behavior is already showing up here.

The play

1. Simulate your customer's search. Put to two or three AI tools the question a real customer would ask: what the best options are for your type of product or service, in your region. Note whether you appear, in what position and with what information. Repeat every month.

2. Make your offer easy to understand, for the machine and for the customer. Spell out, clearly and consistently across all channels, what you sell, to whom, in what price range, with what lead time and with what proof. McKinsey itself cites an account from an e-commerce platform, which sells this kind of solution and did not publish the data, that AI searches finding well-organized catalog data convert about twice as often as those relying on scraped or outdated data.

3. Give the topic an owner. The research recommends putting a senior leader in charge, with real authority. In practice, someone who answers for the company's presence in AI answers and presses marketing, sales and product at the same time.

The other side

The research surveyed consumers, which calls for care in applying these numbers to business-to-business sales. In B2B sales, with several people involved in the decision, contracts and greater risk, the weight of AI is likely to be different, and the research does not measure that. Applying the numbers directly to this type of sale would be an overreach.

The data is also self-reported. Saying AI influenced a purchase is not the same as proving it made the decision. And the comparison of completed purchases over time applies only to the United States.

Finally, the most eye-catching part, the machine buying on its own, is still a minority. Between 21% and 30% of AI users, across the six main purchase categories, say they would let the tool complete a purchase. The shift already weighing on decisions comes earlier, in who makes the shortlist.