What ChatGPT Says About Your Brand Could Cost You the Sale

Your next customer may never see your ad, your window display or your Instagram. They’ll ask a chatbot what to buy, and the chatbot will answer. A New York company called Bluefish is betting big brands will pay to know what it says.

The salesperson you never hired

Picture the question: “What’s the best weekend bag for someone who travels to the Hamptons every Friday?” A few years ago, that question went to Google, a friend, or the man behind the counter at a Madison Avenue boutique. Increasingly it goes to ChatGPT, Gemini or Amazon’s shopping assistant. The assistant names three brands. Yours is either one of them or it isn’t.

That is the problem Bluefish was built to solve. Founded in New York in 2024 by Alex Sherman, Jing Feng and Andrei Dunca, it sells software that shows Fortune 500 brands how AI assistants describe them, and helps them change the answer. Its customers include Adidas, American Express, Hearst and Ulta Beauty. In April 2026 it raised a $43 million Series B co-led by Threshold Ventures and NEA, bringing its total funding to $68 million.

Sherman has done this before. He co-founded PromoteIQ, a retail advertising company Microsoft bought in 2019. His pitch now is that AI is the next major marketing channel, following search, social media and mobile.

Are shoppers really buying through AI?

Yes, and faster than most brands realize. Adobe, which tracks traffic to thousands of US retail websites, reported that visits referred by AI tools rose 393% in the first quarter of 2026 compared with a year earlier.

The more important number is quality. In March 2025, shoppers arriving from AI tools converted 38% worse than other visitors. By March 2026 they converted 42% better, spent 48% longer on the site, and were worth 37% more per visit. In one year, the AI shopper went from window-browser to the best customer in the store.

That makes sense when you think about it. Someone who asks a chatbot for a recommendation and then clicks has already narrowed the field. They aren’t browsing. They’re deciding.

How does ChatGPT decide which brands to recommend?

AI assistants build their answers from what they have read and, increasingly, from what they find when they search the web in the moment. That means a brand’s reputation in AI is shaped by the same things that shape it among well-informed people: clear facts on its own website, reviews, and what credible publications have written about it.

The problem Adobe found is that many brands make that hard. About 34% of retail product pages it examined couldn’t be properly accessed by AI tools, and about a quarter of homepage and category content wasn’t set up for them. If a chatbot can’t read your page, it will describe you using whatever else it can find, or recommend someone else.

Bluefish’s approach, according to Sherman’s comments to Adweek, is not to trick the models. It’s to give them accurate, consistent information and track how the answers change. That is less glamorous than “hacking ChatGPT,” and much more durable.

Why should luxury brands care most?

Luxury has always sold on story: heritage, craft, the right people wearing it in the right places. A chatbot flattens all of that into a sentence. If the sentence is accurate and flattering, it’s the most efficient salesperson you’ve ever had. If it’s outdated, generic or wrong, it quietly sends your customer down the street.

Some brands are moving early. Ralph Lauren launched an AI stylist, Ask Ralph, in its app in 2025, and has said styling and outfit requests make up more than half of how customers use it. The lesson is that affluent shoppers are comfortable asking AI for taste-level advice, not just prices.

What happens this holiday season?

The holidays will be the biggest test yet. Adobe projects US online holiday spending of $275.1 billion from November 1 through December 31, 2026, up 6.7% from last year. It expects traffic to retail sites from AI tools to grow 130% over last holiday season, with the biggest jumps on Thanksgiving (159%), Black Friday (95%) and Cyber Monday (88%).

Adobe’s survey of 5,000 online shoppers suggests why brands should care about more than the traffic. Among shoppers who used AI to help them buy, 77% said they felt more confident in their purchases and 69% said they were less likely to return items. For a luxury brand, fewer returns may matter as much as more sales. A customer who arrives already sure about the bag, the size and the color is the most profitable customer you can have.

Adobe’s earlier data points the same way. AI-referred visitors to retail sites were 12% more engaged than other visitors, stayed 48% longer and viewed 13% more pages. These are people doing their homework, with a chatbot as their research assistant. The brands that show up clearly in those conversations this December will learn quickly how much the channel is worth.

Five things every brand should do now

• • Ask the chatbots yourself. Type the questions your customers ask into ChatGPT, Gemini and Perplexity. Note whether you appear, how you are described, and who you lose to.
  • Make your own website readable. Clear product descriptions, prices, materials and facts in plain text, not just images. If a machine can’t read it, it can’t repeat it.
  • Keep your facts consistent everywhere. Founding date, signature products, where you are sold. Contradictions make AI less confident about recommending you.
  • Earn coverage in publications AI trusts. AI search tools cite articles. Independent press about your brand is one of the inputs that shapes the answer.
  • Measure it every month. AI answers change as the models and the web change. Treat it like search rankings, not a one-time project.

A simple test you can run tonight: ask three assistants the five questions your best customers would ask, such as “best Hamptons caterer for a 60-person dinner” or “most reliable wealth manager for a founder who just sold a company.” Write down every brand they name. If yours isn’t there, or a competitor’s is described more accurately, you have found your first marketing project for 2027.

What this means for New York’s AI gold rush

Bluefish hasn’t disclosed its revenue, which is why it sits in the “watch this space” column of our New York AI Revenue Tracker. But its timing is hard to argue with. The shift Adobe measured, from AI shoppers as an afterthought to AI shoppers as the best customers, happened in twelve months. Brands that wait another twelve will be playing catch-up against competitors who already know what the chatbots say about them.

For the founders building these tools, it’s also a reminder of how New York’s AI companies make money: by selling to industries that already spend heavily, in this case marketing. Whether that becomes profit is a separate question, one we tackle in Growth Is Easy. Here’s Who’s Actually Making a Profit.

The good news for any brand reading this is that the playbook is simple and still early. The weekend bag question is being asked tonight. There is still time to be one of the three names in the answer. And if you know a New Yorker under 40 building the tools behind that answer, nominate them for the Hamptons 40 Under 40.

Frequently asked questions

Sources

• TechCrunch: AI traffic to US retailers rose 393% in Q1 (Adobe data)
• VC Access Online: Bluefish Series B
• Adweek: Bluefish raises $43M
• CMSWire: Bluefish Series A and founders
• Nasdaq (Zacks): Ralph Lauren and Ask Ralph
• Digital Commerce 360: Adobe projects $275 billion in 2026 holiday sales