Zeta Global said on October 5, 2026 that it had signed a definitive agreement to acquire Senso, a Toronto-based, Y Combinator-backed AI engineering company whose technology already helps power Zeta's generative engine optimization product. No price, structure or closing date was disclosed. The deal brings in-house a supplier that Zeta never named when it put the product on sale in September 2025.

In Short

Zeta Global, a large US marketing data company, has agreed to buy Senso, a small startup that helps companies see and change how chatbots such as ChatGPT and Gemini describe their brands. Zeta already sells a tool that does this, and Senso's technology was part of what made it work, so Zeta is buying a company it was already relying on. Nothing changes straight away for the companies that use Senso, since Zeta says it will keep things running for them while the deal goes through.

What Zeta disclosed

The announcement, datelined New York and released at 4:00 PM Eastern Daylight Time, runs to a few hundred words before the legal boilerplate. According to Zeta Global (NYSE: ZETA), the company "has entered into a definitive agreement to acquire Senso, a Y Combinator-backed AI engineering company whose technology helps power Zeta's Generative Engine Optimization (GEO) offering." The deal, according to Zeta, "will bring additional technology in-house, add AI engineering talent and deepen Zeta's connectivity to frontier models."

The financial terms are absent. There is no purchase price, no split between cash and stock, no earnout, no revenue figure for Senso and no expected closing date. Zeta does not say whether regulatory approval is required. Saroop Bharwani, Senso's co-founder and chief executive, "and the Senso team, will join Zeta," according to the company, but the release gives no headcount.

Two procedural details are disclosed. LUMA Partners, the media and technology investment bank, "was an advisor to Senso for this transaction," according to Zeta, which names no advisor of its own. And Zeta "will work closely with the Senso team to provide continuity for customers" - a sentence that confirms Senso has customers beyond Zeta, without naming any of them.

The silence on price fits a pattern in Zeta's recent smaller deals, though not its larger ones. When it agreed to buy LiveIntent in October 2024, Zeta set out a $250 million package of $77.5 million in cash and $172.5 million in stock, with an earnout of up to $75 million over three years. Its agreement for Marigold's enterprise software business, disclosed on September 30, 2025, was priced at up to $325 million. Both were large enough to require detailed disclosure. Senso, by any public measure, is not.

A supplier Zeta never named

The most consequential phrase in the release is a single clause: Senso's technology "helps power" Zeta's GEO product. That relationship had not been part of the public record.

Zeta put its GEO product on sale on September 17, 2025, presenting it as a module inside the Zeta Marketing Platform. PPC Land's coverage of that launch described an "AI Visibility Leaderboard" that tracked brand visibility and sentiment across ChatGPT, Gemini and Claude, flagged citation gaps and off-brand answers, and benchmarked a brand's AI share of voice against competitors. The announcement credited no partner or vendor. Zeta presented the product as its own, built on its proprietary data.

By the second quarter of 2026, GEO had moved into a broader product line. On its earnings call, Zeta placed GEO inside a new Zeta Business Intelligence offering and described API integrations with Claude, ChatGPT and Gemini that let brands monitor and adjust their visibility from one interface. The company told investors this was additive revenue, since it had never sold traditional search engine optimization.

Read against that history, the acquisition is less a new capability than a consolidation of an existing dependency. Whatever share of the GEO product Senso supplied - the release does not quantify it - Zeta will now own rather than license. That has practical consequences for Zeta's margins and for control of the roadmap. It also raises a question for Zeta's clients that the release does not address: how much of the "proprietary" product they were buying was built by a third party.

What Senso is

The release describes Senso briefly. According to Zeta, Bharwani "founded Senso as an AI research lab during the early wave of enterprise deep learning," and brings "more than 20 years of experience working with regulated enterprises." Senso, the release says, "combines applied research and engineering to help businesses turn trusted information into better AI interactions and customer experiences."

Y Combinator's company directory fills in some of the gaps. It lists Senso as part of the Winter 2024 batch, based in Toronto, Canada, with Bharwani and Thomas Nelson as founders and a team size of eight. The directory describes a product that lets companies register agent-first domains, ingest first-party enterprise data and publish verified, structured information that AI agents can use to answer questions and complete transactions. According to the same listing, the platform also tracks how models use that information and adjusts it to improve a company's standing across major large language models. The team size in a startup directory is self-reported and may be out of date; Zeta did not confirm it.

That description points at two distinct jobs. One is measurement: querying AI systems at scale and recording what they say about a brand. The other is supply: publishing structured, verified information in a form that AI systems are likely to retrieve and repeat. The first is what most GEO dashboards sell. The second is closer to what Bharwani emphasised in his statement.

"Senso was founded on the simple principle of making AI more useful in the real world, not more conversational," Bharwani said. "AI discovery is becoming an increasingly important part of the customer journey, but visibility alone is not enough. Together with Zeta, we can combine Senso's AI capabilities with Zeta's proprietary data and customer intelligence to help businesses understand who AI is reaching, improve how they appear in AI-generated answers and connect those insights to measurable outcomes."

The phrase "visibility alone is not enough" is a quiet admission about the category. Tracking where a brand appears in chatbot answers has become a crowded and relatively cheap service. The commercial argument for Senso inside Zeta rests on linking that visibility to customer data and to sales.

From visibility to "who AI is reaching"

According to Zeta, Senso combined with the Zeta Data Cloud "will accelerate Zeta's solutions that help enterprises optimize and remediate how brands reach customers across leading AI platforms, including ChatGPT, Gemini, Claude and Google AI Overviews." The list is one platform longer than the September 2025 launch, which named only ChatGPT, Gemini and Claude. Google's AI Overviews, the summaries placed above conventional search results, are now in scope.

The word "remediate" matters. It implies correcting what AI systems say, not only measuring it. That ties to a recognised problem with these systems: a hallucination, in which a model produces fluent but wrong or unsupported statements, can misstate a brand's prices, products or policies. Zeta's 2025 launch materials listed hallucinations and off-brand answers among the issues the GEO product would flag.

The more ambitious claim is the next one. According to Zeta, the combined offering "will also identify who AI-generated answers are reaching and connect those insights to customer intelligence, marketing activation and measurable business outcomes." Neither Zeta nor Senso explained how.

That is a substantive technical question. AI assistants do not share user-level data about who asked a question or which answer they saw. Referral data is incomplete even at the aggregate level. When Microsoft made its Clarity Citations dashboard generally available in May 2026, the company's own guidance acknowledged that traffic from Google's AI Overviews cannot be reliably separated from standard search, because both share the same referrer. Clarity could identify visitors arriving from ChatGPT, Claude, Gemini and Perplexity, but only once they clicked through to a site.

Zeta's likely route, inferred from its business model rather than stated in the release, is to match whatever traffic and conversion signals it can observe against the Zeta Data Cloud, which the company describes as "one of the industry's largest proprietary data assets." In that model, the identity of the person reached would be estimated after the fact from visits and conversions, not observed at the moment an answer is generated. How accurate that would be, and what consent basis would govern it, are questions the announcement leaves open.

Steinberg's framing, and what came next

David A. Steinberg, Zeta's co-founder, chairman and chief executive, placed the deal within the company's newer description of itself as "the intelligent AI infrastructure company."

"Models are increasingly ubiquitous, but what creates an advantage is the infrastructure that connects an enterprise's data and context to those models, turns the result into action, and turns action into business impact. That is where Zeta plays," Steinberg said. "Senso will bring AI engineering talent and deeper connectivity to the frontier models and personal agents shaping how customers discover brands. The acquisition is an important move in its own right and separate from the announcements we will make at Zeta Live 2026."

Two elements stand out. The first is "personal agents." The phrase reflects a shift that has accelerated since Meta made Muse available in the United States on September 8, 2026, as an AI agent that can browse, fill in forms and complete purchases on a user's behalf. When an agent rather than a person reads the answer and acts on it, the question of how a brand appears in that answer moves closer to the point of sale.

The second is the insistence that the deal was "separate" from Zeta Live, the company's annual conference held in New York on October 8. The distinction did not last long in practice. On October 6, a day after the Senso release, Zeta disclosed an agreement to acquire Digital Audience, a European identity and activation company, again without a price. At Zeta Live itself, Zeta's newsroom lists announcements of AthenaOS, which Zeta calls an enterprise intelligence operating system, and of specialised Athena inference models built with Fireworks on NVIDIA's Nemotron models. Two acquisition agreements and a new platform in four days amount to a coordinated programme, whatever the release says about separation.

Bharwani was scheduled to speak on the Athena Stage at Zeta Live on Thursday, October 8, 2026, "about how GEO is reshaping how brands are discovered, evaluated, and chosen," according to Zeta, with Senso demonstrations in the conference's Product Experience area.

A category moving from startups to owners

The deal comes as measurement of AI visibility shifts from independent startups into larger marketing suites. The clearest marker remains Adobe's agreement in November 2025 to buy Semrush for $1.9 billion in cash, at $12 a share, a deal in which Semrush's generative engine optimization tools were part of the stated rationale. A month later, impact.com made a strategic investment in Evertune, a New York GEO platform founded in 2024 that had raised $19 million, and began integrating its AI brand monitoring reports into its partnership platform.

Demand for these tools rests on an uncomfortable measurement gap. An Adobe survey of more than 500 marketers, published in April 2026, found that 98% lacked a clear, documented AI optimization roadmap and full confidence in their approach, and that 74% either had no measurable strategy for AI search or did not know whether their organisation had one. In its AI Visibility Index published in June 2026, Semrush found that 45% of 481 marketers surveyed said they could not properly measure AI visibility, while only 9% said they could measure every metric they considered relevant. Both are vendor surveys from companies that sell measurement tools. They nonetheless show why a data company sees an opening.

Volatility makes the problem harder. The same Semrush study, built on 126 million US prompts collected from January to April 2026, found that only 36 brands stayed in the top 100 most-mentioned list on every platform in every month. ChatGPT averaged 15.4 cited sources per response, compared with 3.3 for Gemini. Different systems retrieve differently, which is why vendors in this area talk about connectivity to specific models rather than a single optimisation technique. PPC Land's explainer on GEO describes the practice as adapting a site's content, access settings and wider online presence so that AI answer systems choose it as a source.

Zeta's angle differs from Adobe's. Adobe bought a search marketing tool with a large user base of practitioners. Zeta is buying a small engineering team whose work already sits inside its product, and it is pitching the combination not to SEO specialists but to the chief marketing officers and enterprise clients it already serves through the Data Cloud.

Zeta's position going into the deal

The acquisition comes from a company growing quickly through both organic sales and purchases. For the second quarter of 2026, Zeta reported revenue of $442.8 million, up 44% from $308.4 million a year earlier, or 28% excluding Marigold's enterprise business. Adjusted EBITDA was $91.7 million, and the company raised full-year revenue guidance to between $1,811 million and $1,824 million. On the same call, it said the roughly 20% of customers that comprehensively adopt its AI tools account for about 70% of revenue, a company-defined ratio that has not been independently audited.

Zeta's AI narrative has been built in stages. Athena by Zeta became generally available to Zeta Marketing Platform customers on March 24, 2026, built on OpenAI models. On June 23, 2026, Zeta and Palantir said the Zeta Data Cloud would be rebuilt on Palantir Foundry, with Steinberg projecting more than $100 million in annual revenue from the partnership in the coming years. Zeta has also published consumer research on AI shopping, including a survey of 2,000 US adults who had recently bought something with AI, which found that 43% of parents would let AI shop within a set budget, compared with 27% of non-parents.

The company's data practices remain under scrutiny. In July 2026, a federal judge in New York refused to dismiss a securities suit centred on Zeta's past claim of more than 240 million opted-in US individuals, sending the case into discovery. That ruling was not a decision on the merits, and Zeta has called the claims meritless. It is relevant here because the Senso deal's most distinctive promise - linking AI answers to identifiable customers - depends on the same data assets at issue in that litigation.

What remains unknown

Several points are left open. Zeta has not disclosed the price, whether stock is involved, or when the deal will close. It has not said how many people are joining, whether the Senso brand will survive, or whether Senso will continue selling to companies that are not Zeta clients. The continuity commitment covers customers, but it does not say for how long, or on what terms.

The release also does not describe the existing commercial arrangement between the two companies: whether Zeta licensed Senso's technology, paid for engineering services, or held an equity stake. Nor does it explain what "deeper connectivity to frontier models" means in practice. Large AI developers offer standard APIs to any paying customer, so the phrase may refer to engineering experience in using them well rather than to any privileged access. Zeta did not say.

Finally, the forward-looking statements section, as is standard, warns that actual results may differ materially from expectations. Until the deal closes and Zeta reports on it, the clearest fact in the announcement is the one it mentions almost in passing: part of the AI search product Zeta has been selling for a year was already being built by the company it now plans to buy.

Timeline

Summary

Who: Zeta Global (NYSE: ZETA), led by co-founder, chairman and chief executive David A. Steinberg, and Senso, a Toronto-based, Y Combinator-backed AI engineering company led by co-founder and chief executive Saroop Bharwani. LUMA Partners advised Senso.

What: Zeta signed a definitive agreement to acquire Senso, whose technology already helps power Zeta's generative engine optimization product. Bharwani and the Senso team will join Zeta. No price, structure or closing date was disclosed.

When: Zeta disclosed the agreement on October 5, 2026, three days before its Zeta Live 2026 conference on October 8, and a day before it disclosed a separate agreement to acquire Digital Audience.

Where: The announcement was datelined New York. Senso is based in Toronto, according to Y Combinator's directory. The product covers ChatGPT, Gemini, Claude and Google AI Overviews.

Why: Zeta says the deal brings technology in-house, adds AI engineers and deepens its connectivity to frontier models, as brands try to measure and correct how AI assistants describe them. For marketers, it marks the absorption of an AI visibility supplier into a large data company, and comes with an untested promise to identify who AI-generated answers reach.