Gong goes from telling you how to win to winning deals for you: 'Activate' lands at Celebrate '26

Gong has spent a decade listening to sales calls and telling reps how to close. Now it wants its AI to do the closing-adjacent work itself. In a press release timed to Celebrate '26, the company's annual customer conference in Las Vegas, Gong announced Gong Activate — "the next generation of Revenue AI," in the company's words — alongside a product bundle called Mission Callisto that adds data enrichment, custom agent building, and a deep-research mode to the platform.

Gong pioneered what it calls Revenue AI back in 2015: models that observe sales conversations and guide reps. Activate is the shift from guidance to action. Always-on agents continuously evaluate every account, opportunity, and revenue play, build detailed prescriptions for the best path forward, and automate the time-consuming steps in alignment with the plan. Outcomes then flow back into the Gong Revenue Graph, sharpening the model for the next decision — a self-improving loop the company says will let new reps perform like seasoned veterans.

"For years, AI told you how to win. Now, AI wins deals for you," said Bendov. "That means faster cycles, higher win rates, and reps running proven plays instead of guessing."

Mission Callisto: the supporting cast#

Activate is the headline, but the Mission Callisto release ships the pieces that make it workable in practice. The biggest is Gong Enrich, which fills the data gaps that agents choke on: it automatically completes missing account and contact details, either with a single click or at scale through Custom Agents triggered on defined conditions. Launch data partners include Apollo, Findymail, Firmable, Kernel, LeadIQ, Lusha, RocketReach, Wiza, ZeroBounce, and ZoomInfo. Because enriched data flows back into the Revenue Graph, the AI's answers get sharper everywhere — assistants, agents, and dashboards alike.

"Every revenue decision is only as good as the context behind it," said Eilon Reshef, Gong's chief product officer and co-founder. "Gong Enrich makes sure that context doesn't stop where the conversation ends."

Agent Builder lets revenue teams build Custom Agents that fire automatically when defined events and conditions occur — a team can specify triggers, conditions, and actions in a visual editor or simply describe the agent in natural language. And Deep Mode, an agent inside Gong Assistant, takes on the investigation behind hard strategic questions, turning hours of manual digging across systems into one grounded answer. The Assistant is also now embedded in Deal Boards and Account Boards, and can build Gong Engage Flows from natural-language requests.

The proof is in the customer math#

Gong brought customers on stage to put numbers behind the pitch. AT&T Business reported a 54% improvement in rep productivity from scorecards, call reviews, and manager feedback inside Gong. Cisco, a year into its rollout, has deployed 18,000 Gong licenses company-wide, extracted insights from more than 450,000 conversations, and run over 50,000 Gong Engage flows — with early-adopting teams showing roughly 32% larger deal sizes and 26% higher win rates. Experian, which consolidated a division built from four acquisitions onto the platform, reported a 25% lift in win rates.

The pattern across these announcements is unmistakable: the enterprise AI conversation has moved from "can the model reason?" to "will the agent do the job, and can we trust it?" Gong's answer is context — a decade of institutional knowledge captured in the Revenue Graph — plus execution that stays inside the playbook. Whether buyers will hand revenue workflows to agents is now the question every revenue AI vendor, including Gong's new partner-turned-customer Anthropic, is racing to answer.