Twilio's AI Story Gets a Balance Sheet

Record Q3 profit and a 2027 framework landed, but the stock is 20% off its October high—Twilio still has to prove its identity and Conversations layer can bend gross margin.

TWLO · Earnings Call · 2026-10-29

Twilio reported the kind of quarter that should end a debate. Revenue was $1.56 billion, up 20% reported and 16% organically. Non-GAAP gross profit rose 18% to a record $768 million. Non-GAAP income from operations hit a record $326 million, up 39%, and free cash flow hit a record $370 million. Management raised full-year 2026 organic growth to 15% and gave a first look at 2027: at least 10% organic growth, at least $1.35 billion in non-GAAP operating income, and at least $1.35 billion in free cash flow. Yet the tape is not celebrating. Twilio enters the print 19.7% below its Oct. 1 high and still 45.5% below the 2021 peak. The market is not disputing the quarter. It is asking what the AI layer is actually worth.

The carrier-fee fog around a clean operating beat

The first thing to price is the gap between reported and organic growth. The difference is largely carrier fees: $53 million of incremental US pass-through fees in Q3, $75 million guided for Q4, and $245 million for the full year. Gross margin fell 90 basis points year over year to 49.2%, but excluding the fees it rose 80 basis points. Aidan Viggiano was explicit: “Without these incremental fees, non-GAAP gross margin would have been up 80 basis points year-over-year and up 30 basis points quarter-over-quarter.” — Aidan Viggiano, Chief Financial Officer · 2026-10-29 This is not a new issue. On the October 2025 call, Aidan described the fees as “a gross up of revenue and a gross up of cost of goods sold, but it has no impact on our ability to generate gross profit dollars” — Aidan Viggiano, Chief Financial Officer · 2025-10-30. What is new is scale: the pass-through is now large enough to obscure the underlying margin improvement and flatter reported revenue growth. The second thing to price is the beat itself. Q3 organic growth beat the high end of guidance by roughly four points, inside the 2-4 point range Aidan called normal. Management explicitly pushed back on the 5%+ beats from the first half. That is a maturity signal: the base business is durable, but it is no longer accelerating at the pace that made the first half feel like an inflection.

Identity is the first genuinely new category in years

The most important new language on the call was identity. Twilio put an identity offering into private beta, aimed at validating who an agent represents across applications, channels, and protocols. Khozema Shipchandler framed the use cases broadly: “The product is meant to suit all of the use cases that you referenced, agent-to-human, human-to-agent, agent-to-agent.” — Khozema Shipchandler, Chief Executive Officer · 2026-10-29 Monetization is not finalized, but management expects usage-based pricing and software-like gross margins, with general availability next year. This is genuinely new. Prior calls treated identity through the Stytch acquisition and Verify; Q3 turns it into a standalone layer for the agent economy. The strategic logic sits in consumer agent fragmentation. Every agent platform needs a neutral validator, and Twilio already sits in the transactional flow. Khozema's framing:

we are the neutral party. And so virtually every player in the space who are predominantly participating at the application layer, they are running off of Twilio's rails.
Khozema Shipchandler, Chief Executive Officer · 2026-10-29

That is the bull case in one line. The adjacent product is the Conversations layer, which landed Twilio's first eight-figure deal with a health insurance provider in Q3. The customer uses Conversation Relay and real-time transcription to power a voice AI agent, cut wrap-up time, and track calls. If identity and Conversations become the software layer on top of messaging and voice, Twilio's margin story changes. If they remain science projects, the multiple does not.

Voice AI is a 1% penetration story—and that is the point

Voice is the proof point that the AI layer is not imaginary. Voice revenue grew more than 20%, software add-ons grew more than 25%, and self-service voice grew more than 50%. But Khozema kept the enthusiasm anchored: “it's still 1% penetration in terms of the total available voice minutes out there.” — Khozema Shipchandler, Chief Executive Officer · 2026-10-29 He has been saying some version of this for quarters. In August 2026 he called it “very early innings, very early innings” — Khozema Shipchandler, Chief Executive Officer · 2026-08-06; in April 2026 he said AI was a catalyst but “not meaningfully contributing to the overall results” — Khozema Shipchandler, Chief Executive Officer · 2026-04-30. The difference now is that management is willing to underwrite a 2027 framework rather than just describe the opportunity. That framework does not require a sudden AI step-function. It requires the base business to keep compounding and the new products to add incremental gross profit. The analysts found the soft spot immediately. Ittai Kidron asked why gross profit should grow only in line with revenue: “why should gross profit grow in line with revenue? Why should it not grow faster?” — Ittai Kidron · 2026-10-29 Aidan's answer was mix. Messaging is about 60% of revenue and structurally lower margin; Q4 also carries holiday hosting costs and the lapping of email cloud-migration tailwinds. That is the central tension. Twilio's AI products are software-margin, but they are not yet big enough to bend the consolidated gross margin. Until identity and Conversations scale, the P&L is still a messaging business with a software attach.

The keyword that vanished: Segment

One of the more telling non-mentions is Segment. In 2024 and 2025, Segment was a top-30 keyword cluster—standalone Segment business, Segment revenue, Segment into Communications. In the current top-30, it is gone. The new console, which dominated the August 2026 call, is also absent. What replaced them is a simpler story: Conversations, identity, branded calling, voice AI. Khozema had already signaled the shift in April 2026: “We are not as focused on Segment as a standalone. We are much more interested in using the data technology to enrich communications.” — Khozema Shipchandler, Chief Executive Officer · 2026-04-30 The Q3 call completes the rhetorical pivot. Segment is now context, not a product line. That is strategically cleaner, but it removes a potential standalone catalyst and makes the AI layer carry more of the narrative weight. The valuation debate is now about what multiple the AI layer deserves. At $37 billion market cap, Twilio trades at 5.7x revenue, up 41% year over year, and 11.7x gross profit, up 45%. That is not 2021 mania—the stock is still 45.5% below its peak—but it is a meaningful re-rating off the lows. The 90-day return of 28.4% and the 19.7% drawdown from the Oct. 1 peak suggest the market has already started to pay for the AI option. Q3 justified the base case; it did not yet justify a higher multiple on its own. The next proof points are concrete: identity general availability and pricing next year, Conversations scaling beyond the first eight-figure deal, voice AI moving from 1% penetration into regulated verticals, and gross profit growth separating from organic growth. Twilio has spent three years proving it can grow and generate cash. The Q3 report says it can. The next year is about proving the AI layer is not just a narrative wrapper around messaging. The stock's October drawdown is the market saying: show me the gross profit.