Behind Zeta Global’s big (and early) AI bet | David Steinberg

Zeta Global is an AI-native marketing platform processing trillions of consumer signals for Fortune 2000 companies. The company went public on the NYSE in June 2021 and is projecting close to $1.8 billion in net revenue for 2026 — on the back of four consecutive years of 30%+ organic compounded top-line growth and a 120% net retention rate. In a recent episode of Unicorn Builders, we sat down with David Steinberg, CEO, Co-Founder, and Chairman of Zeta Global, to understand how a seven-time founder walked into a 2017 board meeting and convinced his board to go from $50M in annual EBITDA to zero for three years to rebuild the entire platform around native AI — at a time when people were still asking, "Who's Al and why is he in charge of my data strategy?"Topics Discussed:The board meeting where David proposed sacrificing $50M+ in annual EBITDA to rebuild from scratchHow Zeta used two acquisitions — Boom Train (AI/NLP) and Disqus (publisher data) — as the architectural foundationWhy running the legacy platform for cash flow was what made the rebuild possibleWhat competitors got wrong: acquiring platforms and freezing the architecture in placeHow ChatGPT turned years of invisibility into a firehose of inbound from Fortune 2000 boardsWhy David distinguishes "aptitude for risk" from both blind risk and no riskGTM Lessons For B2B Founders: Sell the floor to the board, not just the ceiling: David didn't win his board by projecting a grand slam. He closed by making the downside concrete — if the new platform failed in year three, they still had the legacy business, most customers, and a profitable exit. Founders pitching transformational bets should lead with what the floor looks like, not just the ceiling.Run the legacy for cash while you rebuild: Zeta didn't shut down its ~$200M revenue business during the re-architecture — "it was the reason we didn't lose money." That cash funded the new engineering team. Founders making major pivots should identify what existing revenue can fund the transition rather than burning runway on both sides.Earn customer ownership before the product ships: While building, Zeta showed customers the roadmap and solicited "dream features" so they could "feel ownership" in what was coming. That investment drove high migration rates at launch. Loop customers into the build early — not for validation, but to make the migration ask easier when the time comes.Native AI architecture compounds; API integrations stagnate: Most competitors bolt AI onto legacy architecture via API — platform to algorithm to data repository and back. That latency destroys ROI at scale. Zeta's data cloud is sole-sourced to its own platform, never exposed to external models — making the intelligence proprietary and compounding. Most competitors made acquisitions a decade ago and are still running the same architecture they bought."AI roadmap AI" is not a moat: After ChatGPT, David says most competitors are running what he calls "AI roadmap AI — where it's not real just yet." Zeta was already on its fourth iteration of its AI agent studio and had launched Athena, a voice-enabled super agent. The post-ChatGPT demand window went to companies that had been building for years before it mattered.// Sponsors: Front Lines — We help B2B tech companies launch, manage, and grow podcasts that drive demand, awareness, and thought leadership. www.FrontLines.ioThe Global Talent Co. — We help tech startups find, vet, hire, pay, and retain amazing marketing talent that costs 50-70% less than the US & Europe. www.GlobalTalent.co//Don't Miss: New Podcast Series — How I Hire Senior GTM leaders share the tactical hiring frameworks they use to build winning revenue teams. Hosted by Andy Mowat, who scaled 4 unicorns from $10M to $100M+ ARR and launched Whispered to help executives find their next role. Subscribe here: https://open.spotify.com/show/53yCHlPfLSMFimtv0riPyM