Skip to main content
Mark Zuckerberg speaks at Meta event, discussing AI's role in driving business results and monetization.

Editorial illustration for Zuckerberg: Meta to get paid when AI delivers business results

Meta's AI to Charge Based on Business Results

Zuckerberg: Meta to get paid when AI delivers business results

4 min read

Mark Zuckerberg told investors on Wednesday that Meta's enterprise AI plans go well past the customer-service agent the company launched in June. On the second-quarter earnings call, the Meta CEO laid out a broader pitch: APIs, business agents, compute sales, and other services built for large customers. The goal is a new revenue line that doesn't depend on advertising, which still generates most of Meta's money, or subscriptions, which bring in far less.

The near-term plan starts with familiar territory. Meta wants to sell AI agents to the advertisers and small businesses already running campaigns on its platforms, letting those companies field customer interactions through messaging apps rather than employees. Zuckerberg framed the pricing model as an extension of how Meta already charges for ads: get paid based on outcomes for the business, not just placement.

Beyond that base, Zuckerberg pointed to a second tier of customers Meta hasn't historically served this way. He described plans to open up internal tools, including systems Meta built for its own coding and development work, to outside enterprise clients. That's where he picked up the thread on what those tools might look like in practice.

In June, Meta entered the enterprise AI market with a new AI agent aimed at businesses, to help with customer service, support, and other daily operations. But the tech giant’s enterprise AI ambitions are much more expansive, Meta CEO Mark Zuckerberg told investors on Wednesday’s second-quarter earnings call.

Why this matters

Zuckerberg is telling us Meta wants to run its enterprise AI business the same way it runs ads: charge on outcomes, not seats or tokens. That's a different pitch than OpenAI's per-seat Copilot model or Salesforce's Agentforce consumption pricing, and it plays to Meta's actual strength, which is billions of dollars in ad infrastructure already built to measure whether a dollar spent produced a result. For founders building on top of Llama or Meta's business agents, the question is what "results" means in a contract and who defines it when a support ticket gets resolved or a lead converts.

Performance-based pricing sounds founder-friendly until you're the one arguing with Meta's attribution model. For researchers, it's a signal that the next fight in enterprise AI isn't just about model quality, it's about who owns the measurement layer. Meta already owns that layer for advertisers.

Whether it can extend the same trust to CFOs evaluating agent ROI is the thing worth watching next.

Common Questions Answered

What enterprise AI services is Meta planning to offer beyond the customer service agent launched in June?

Meta plans to offer APIs, business agents, compute sales, and other services built for large customers. These offerings represent a broader enterprise AI strategy that extends well beyond the initial customer service agent, creating a new revenue line independent of advertising and subscriptions.

How does Meta's outcome-based pricing model for enterprise AI differ from competitors like OpenAI and Salesforce?

Meta plans to charge customers based on business results and outcomes rather than per-seat licensing or token consumption, similar to its advertising model. This approach differs from OpenAI's per-seat Copilot model and Salesforce's Agentforce consumption pricing, leveraging Meta's existing infrastructure for measuring return on investment.

Why is Meta's existing ad infrastructure an advantage for its enterprise AI business model?

Meta has already built billions of dollars in ad infrastructure designed to measure whether money spent produces measurable results and outcomes. This existing capability gives Meta a competitive advantage in implementing outcome-based pricing for enterprise AI services, as the company already possesses the tools to track and verify business results.

What is Meta's goal with its new enterprise AI revenue line?

Meta aims to create a new revenue stream that doesn't depend on advertising, which currently generates most of the company's revenue, or subscriptions, which bring in far less. This diversification strategy reflects Meta's ambition to reduce reliance on its traditional advertising business model.

LIVE01:06Zuckerberg Says Personal AI Agents Will Drive Meta's Next Products