OpenAI’s Agent Revenue Miss: The Growth Bet That Slipped

OpenAI’s original thesis depended heavily on agents driving the next wave of revenue. The Q3 2025 reforecast tells a different story.

The Revenue Miss

• Agent revenue cut from $2.8B to $1.4B—a 50% reduction
• Internal benchmarks weren’t met
• December 2025 “code red” memo delayed agent initiatives to focus on Gemini 3 competition

Agentic Commerce Reality

OpenAI has launched the Agentic Commerce Protocol with Stripe, with partnerships spanning Shopify, Etsy, Walmart, Target, Instacart, and DoorDash—over 1 million merchants can theoretically reach 800M weekly ChatGPT — as explored in the intelligence factory race between AI labs — users.

But as one industry observer noted: “Everyone is building the spaceship, but no one has really launched it yet.”

The Opportunity vs. Reality Gap

• McKinsey TAM: $3-5 trillion in global retail by 2030
• Current e-commerce sessions: 0.82% of total referrals
• Protocol fragmentation: OpenAI’s ACP vs Google’s AP2

The problem isn’t that agentic commerce won’t matter. The problem is OpenAI projected revenue arriving faster than the market validated, forcing reliance on other streams to fill the gap.

This is part of a comprehensive analysis. Read the full analysis on The Business Engineer.