Zeta Global at Goldman Sachs Communacopia + Technology Conference 2026: ai push

Zeta reported strong latest-quarter growth, with revenue up 44% year over year, EBITDA up 56%, and free cash flow up 73%, alongside a Rule of 40 metric of 67-68. Athena adoption reached 40% of super-scaled customers within four months, with 83% of engagement occurring via voice, while pipeline growth accelerated to 60% year over year in Q2 excluding Palantir-related opportunities and closed deal sizes rose 40%. Management maintained that guidance embeds 2-5 points of conservatism, citing net revenue retention above its 110%-115% target, average customer spend growth of about 20%, and expanding distribution through Palantir, Snowflake, OpenAI and the Gap win. Shares rose 1.1% after hours to $31.13, although the stock remains valued at high EBITDA and EBIT multiples.

Analysis

ZETA’s investable differentiation is not the AI interface but the conversion of higher platform engagement into usage-based revenue, while development costs appear structurally contained. If agent-led workflows lift campaign creation and activation frequency, incremental revenue should carry materially higher contribution margins than the legacy sales model; the key proof point over the next 1-3 quarters is sustained expansion in consumption revenue and NRR rather than reported AI-user adoption. This creates a credible multiple-expansion case if ZETA can demonstrate that AI is an ARPU accelerator rather than a retention feature bundled into existing contracts.

The second-order competitive pressure falls most directly on ADBE, CRM and ORCL, whose marketing products are often embedded in broader workflow suites and may face greater pricing pressure during stack replacements. SNOW and PLTR benefit from faster enterprise data onboarding and co-sell validation, but the economics for each partner are likely immaterial near term; PLTR’s upside is primarily a pipeline/credibility halo rather than a revenue estimate change until joint wins are disclosed. Agency-channel expansion can improve sales efficiency but also raises cyclicality: advertising budgets are among the first enterprise spend pools cut if elevated rates weaken consumer demand.

Consensus may over-credit the data moat and under-price execution, privacy, and autonomy risk. The claimed accuracy, ROI and fixed-cost inference economics are management assertions; a single material campaign error, consent/data-governance challenge, or slower conversion from exploration to autonomous activation could compress a premium growth multiple quickly. Near-term upside catalyst is the product event and subsequent guidance revision; 6-18 month upside requires evidence that larger enterprise deployments translate into durable margin expansion, not merely a conference-driven narrative.

Key Decisions for Investors

• Initiate a starter long ZETA only on post-event confirmation: add if shares hold above the prior 52-week high (~$32.81) for 3-5 sessions and management raises full-year revenue or EBITDA outlook. Target 20-30% upside over 6-12 months from sustained consumption/NRR evidence; exit if NRR falls below the 110% floor or EBITDA growth decelerates below revenue growth for two quarters.
• Use a relative-value expression: long ZETA / short ADBE in equal-dollar size over 3-6 months. The thesis is that replacement-cycle share gains and usage monetization can re-rate ZETA while Adobe faces AI-driven feature commoditization in marketing workflows; stop the spread if ZETA fails to convert pipeline into bookings or Adobe shows accelerating Digital Experience growth.
• Do not chase PLTR on the partnership narrative. Set an alert for disclosed joint contract value, named enterprise wins, or PLTR commercial revenue guidance attributable to the relationship; absent those datapoints, the likely financial contribution is too small to justify incremental PLTR exposure at a premium valuation.
• Watch ZETA’s next earnings for consumption revenue mix, gross margin, customer implementation duration, and autonomous-activation penetration. A rise in token/inference cost above management’s implied de minimis level, or increased services staffing required to supervise deployments, would invalidate the operating-leverage thesis and favor reducing exposure.