ZETA: Is AI Turning Marketing Software Into an Intelligence Platform?

First, what does Zeta actually do?

Large companies know a lot about their customers, but that information is often fragmented across CRM systems, websites, apps, email platforms, advertising systems and other databases. Zeta's pitch is to connect those signals, resolve identity, turn them into intelligence and then activate marketing from the same platform.

The potential advantage isn't simply an AI interface. It is the combination of data, identity, decisioning and activation in one workflow. That is what could move Zeta from being another marketing tool toward something customers rely on more deeply.

Athena may be the most important part of the story.

Athena is Zeta's agentic AI layer for enterprise marketing. Rather than forcing a marketer to navigate dashboards, the goal is to turn questions into answers and answers into actions: identify an audience, understand an opportunity, build a campaign, activate it and measure the outcome.

Zeta announced a strategic collaboration with OpenAI in January 2026 to support Athena's conversational intelligence and agentic applications. The early usage signals are encouraging, but I don't want to confuse usage with monetization. The real test is whether Athena increases retention, customer spend and ultimately cash generation.

OpenAI, Palantir and the enterprise AI stack.

The partnerships are worth understanding, but not all logos mean the same thing. OpenAI supports the model and conversational layer. Zeta and Palantir announced a deeper infrastructure relationship in June: Zeta said its Data Cloud will be rearchitected on Palantir Foundry, connecting customer intelligence with enterprise operational data.

The important question isn't whether Zeta has impressive partners. It's whether those relationships make Zeta more embedded in enterprise workflows and harder to replace.

The numbers are beginning to look different.

Zeta produced $1.305 billion of revenue in 2025, up 30%. In Q2 2026, revenue reached $443 million, up 44%, while adjusted EBITDA reached $92 million and free cash flow reached $58 million. Management raised its 2026 revenue guidance midpoint to $1.818 billion.

What I like here is that the story is moving beyond revenue growth alone. Cash generation is scaling and GAAP profitability is beginning to appear. That is the type of financial inflection that can change how the market values a software company.

But 44% growth needs context.

Zeta has been acquisitive. LiveIntent and Marigold's enterprise business added capabilities and revenue, which means headline growth isn't the same as underlying growth. Management's current full-year outlook is roughly 39% reported growth versus about 24–25% when excluding M&A and political-candidate revenue.

The customer data may be even more interesting.

Super-scaled customers—customers generating at least $1 million of trailing-12-month revenue—rose from 148 at the end of 2024 to 184 at the end of 2025 and 197 by Q2 2026. Zeta's annual NRR also increased from 113.6% in 2024 to 128.0% in 2025.

There is another detail I like: 90 of Zeta's 184 super-scaled customers at year-end 2025 had been customers for at least five years, and those long-tenured relationships represented 68.2% of super-scaled customer revenue. This isn't purely a new-logo growth story.

There may be a flywheel here. But I want proof.

Having data does not automatically create a moat. Salesforce, Adobe and large enterprises themselves also possess enormous amounts of data. Foundation models are improving quickly. The real moat test is whether Zeta's identity resolution, proprietary signals, workflow integration and measurable outcomes make the platform increasingly painful to replace.

Zeta is already pushing beyond traditional search.

On October 5, Zeta announced an agreement to acquire Senso, a Y Combinator-backed AI engineering company. Senso's technology supports Zeta's generative-engine-optimization offering, aimed at helping brands understand and improve how they appear across AI-driven discovery experiences.

I wouldn't make Senso a major part of the thesis yet—the transaction is new. But strategically it fits the direction Zeta is signaling: follow consumer discovery as it shifts from traditional search toward AI interfaces, and connect those interactions back into Zeta's customer-intelligence stack.

What could break the thesis?

The business may be at an inflection. The stock already knows it.

Zeta's improving growth, cash generation and AI positioning have attracted more investor attention. That means I don't want the article to end with a simplistic “great company = buy.” As expectations rise, execution matters more.

I care especially about free cash flow per share. Stock-based compensation and dilution can coexist with strong reported free cash flow, so the shareholder-level outcome matters more than the headline cash number alone.

What would have to happen from here?

The numbers I want to keep watching.

Primary research