Twilio's Q4 Slowdown Tests the AI Infrastructure Thesis

Twilio's third quarter delivered more growth and cash than Wall Street expected, yet the stock fell more than 16% because the next quarter points to a clear slowdown. That contrast matters more than the headline earnings beat: Twilio is improving its economics while investors debate how quickly artificial intelligence can extend the growth cycle.

Revenue rose 20% from a year earlier to $1.56 billion, according to Twilio's October 29 results. Organic revenue, which removes acquisitions, divestitures and certain carrier-fee effects, grew 16%. The company reported $148.6 million of operating income under generally accepted accounting principles and $325.7 million on an adjusted basis. Free cash flow increased to $369.6 million from $247.5 million a year earlier, while the dollar-based net expansion rate improved to 114% from 109%.

Those figures show that existing customers are spending more and that Twilio is converting more of its growth into cash. Management raised its full-year outlook to 20% reported revenue growth and 15% organic growth. It also increased expected adjusted operating income and free cash flow to $1.19 billion to $1.20 billion each. Twilio repurchased $127.2 million of stock during the quarter and had about $700 million remaining under its current authorization, adding capital returns to the improved cash profile.

The market's concern is the fourth quarter. Twilio forecast revenue of $1.595 billion to $1.605 billion, equal to 17% to 17.5% reported growth, but only 11% to 12% organic growth. Adjusted operating income is expected to fall sequentially to $300 million to $310 million. For a stock that had climbed sharply before the report, that deceleration left little room for disappointment.

Chief Executive Khozema Shipchandler argued that AI agents will eventually require more voice, messaging and email infrastructure. He estimated that voice AI represents only about 1% of customer voice interactions and said Twilio can supply the communications layer rather than compete with every application developer. That is a plausible strategic advantage because usage can rise across many AI products without Twilio having to pick the winning consumer app.

It is not yet a near-term financial promise. Shipchandler said businesses still block some automated agents, regulated industries are likely to adopt later, and Twilio's agent identity product remains in private beta. He does not expect that product to become a meaningful revenue contributor in 2027. Twilio's usage-based model also makes demand harder to forecast because revenue depends on customer activity rather than fixed subscription commitments.

The investment question is therefore about timing, not whether Twilio has exposure to AI. The current business is producing stronger margins, cash flow and expansion from existing customers. The share-price reaction says investors want proof that those improvements can persist as organic growth slows, and that AI demand can move from an attractive infrastructure story to measurable usage before expectations run ahead of revenue.