He spotted retail media before it had a name. Now he is teaching the Fortune 500 how to talk to a machine that answers before anyone clicks.
The guy who reads the room before the room knows it's a room.
Type a question into Google now and a synthesized answer arrives before your eyes reach the first link. Alex Sherman noticed his own hand stop reaching. "There's almost a new muscle memory where I don't quite want to go poking through all of those links," he says. That small, private hesitation became the thesis for a company.
Bluefish, which Sherman runs as CEO and co-founded in 2024, measures and shapes how brands appear inside generative AI answers - ChatGPT, Google's Gemini, Anthropic's Claude, Perplexity, and Amazon's Rufus. Not the ads beside the answer. The answer itself. When a shopper asks an assistant which running shoe to buy, or which credit card to carry, the model's reply is the new storefront window. Sherman sells the Fortune 500 a way to see through it.
The platform processes millions of AI prompts and responses every day, across more than a dozen verticals - financial services, pharmaceuticals, beauty, luxury, consumer packaged goods. The job is not vanity tracking. "We help clients understand not just how they are performing but why," Sherman says, "and which pieces of content, which creators, and which publications are shaping the models' understanding of their business." It is forensic work dressed as marketing.
Having reached over 1 billion monthly users within 12 months of launch, AI is clearly the next major marketing channel on the internet - just like search, social or mobile before it. Alex Sherman, on why the timing is not optional
The category has a name now - generative engine optimization, or GEO - and Sherman is careful not to let it sound small. He does not frame Bluefish as a smarter SEO tool waiting for the next algorithm update. "We think pretty holistically about GEO beyond just the next chapter of search," he says. "We think of it as a fundamental rebuilding of the mar tech stack." That is the difference between selling a flashlight and selling the wiring.
Start with the detail that does not fit the resume: Sherman studied European history at Columbia. No computer science, no MBA pipeline. In 2008 he became one of the earliest employees at MediaMath, the demand-side platform that helped invent programmatic advertising. He learned the trade by being in the room while the room was still being built.
In 2012 he and Peter Schwartz, another MediaMath alum, started a company called Spotfront. It grew up into PromoteIQ - a platform that let retailers sell advertising to the brands on their shelves. Today the industry calls that "retail media" and analysts measure it in the tens of billions. In 2012 it did not have a tidy name. Microsoft acquired PromoteIQ in 2019. The company would later be wound down, the way acquisitions often quietly are. Sherman does not pretend otherwise.
What links MediaMath, PromoteIQ, and Bluefish is not a vertical or a technology. It is timing. Each was a bet placed before the consensus formed - programmatic, then retail media, now AI answers. Sherman keeps finding the channel while it is still a hunch, then builds the instrument that lets enterprises measure it.
It is rare to have a chance to redesign how an entire industry works. Alex Sherman, on the appeal of building Bluefish
For Bluefish he brought in two co-founders who also know what it feels like to build something a giant later buys. Andrei Dunca founded LiveRail, acquired by Facebook. Jing Feng rounds out a trio that, between them, has built and sold category-defining platforms now owned by Microsoft and Meta. They are not first-timers wandering toward a trend. They are repeat operators who recognize the shape of one.
Sherman's case is blunt. "What worked for the old internet will not work for the next version. AI marketing is essential." If you accept that LLMs are trained on the open web, the conclusion follows fast: "The team that's writing copy for your brand.com or for your retailer product detail page needs to start to optimize and tune that content for LLMs." The audience for your words is no longer only a human. It is also the model that will summarize you to a human who never sees your words at all.
That reframing is the quiet violence in his pitch. Two decades of marketing wisdom assumed a person on the other end - someone to persuade, to retarget, to nudge toward a click. Bluefish assumes a machine reads first and decides what the person hears. The brand's job shifts from winning attention to influencing a summary. Sherman has spent his whole career selling enterprises the dashboard for a channel they did not yet know they had to manage. He is doing it for a third time, and this time the channel reached a billion people before the marketing departments finished their slide decks.
The temptation with any new channel is to sell fear. A brand discovers that ChatGPT describes its flagship product inaccurately, panics, and pays someone to make the bad answer go away. Sherman is building for the opposite posture. Bluefish frames itself as enterprise-grade infrastructure - the same category of tooling a Fortune 500 marketing org already runs for search, social and programmatic - rather than a one-off cleanup service. That is why the language matters when he says brands want "the same enterprise-grade sophistication that they expect across their existing marketing stack." The bet is that AI visibility becomes a permanent line item, governed and measured, not a fire to put out once.
The early access helped. Bluefish went through Comcast NBCUniversal's LIFT Labs accelerator, and Sherman is candid about what that bought a young company: a fast track to scale. "Within days, we met with leaders from three major divisions," he has said. "That level of access would have taken months or years outside of the program." For a platform whose entire value proposition depends on selling to enormous, slow-moving organizations, shaving years off the first conversations is not a perk. It is close to a business model.
What Sherman is selling, underneath the acronyms, is a way to stay legible in a world that increasingly answers on your behalf. He has watched this movie before - twice. The plot does not change much: a new layer appears between the brand and the buyer, nobody quite knows how to measure it, and the company that builds the ruler ends up indispensable. The only question that interests him now is whether the AI internet is the biggest layer yet. His billion-user number suggests he already has his answer.
Bluefish counts some of the most brand-obsessed companies on earth among its customers. These are organizations that spend lifetimes controlling how they look in a magazine spread - and now have to worry about how they read in a chatbot's reply.
The April 2026 Series B drew a telling roster of backers: Threshold Ventures and NEA co-leading, with Amex Ventures, TIAA Ventures, Salesforce Ventures and Bloomberg Beta alongside. When enterprise software and financial heavyweights write checks into your marketing tool, they are not buying a trend. They are hedging their own visibility.
"There's a sort of new expectation of an instantaneous synthesized answer, and I don't think I'm alone in that experience."
"Enterprise brands are looking for agentic marketing technology partners with the same enterprise-grade sophistication that they expect across their existing marketing stack."
"We want to be the enterprise marketing platform for AI, helping the world's largest brands."
"If LLMs are trained on the internet, then the team writing copy for your brand.com needs to start to optimize and tune that content for LLMs."
Sherman makes the rounds explaining agentic AI to people who sell things for a living. A good place to start is his Black Friday conversation on the AI takeover of e-commerce, where he walks through what happens when shoppers ask an assistant instead of a search bar.
Black Friday's AI Takeover: Bluefish CEO Alex Sherman on Agentic AI →
For longer form, he sat down with AdExchanger Talks on "Marketing In The Age Of AI Answers," and with the RiskReversal podcast on agentic AI's disruption of e-commerce.
Sources: PR Newswire, Adweek, AdExchanger, Comcast NBCUniversal LIFT Labs, Crunchbase, AlleyWatch, FinSMEs, citybiz. Facts drawn from public reporting and interviews; figures reflect reporting as of mid-2026.