Twilio: AI Growth Is Real, But the Rally Already Priced It In

Twilio's stock has nearly doubled over the past year to roughly $240, and the crowd's story fits in a sentence: AI is finally pulling the company's growth back up. The operating numbers say that story is more than marketing this time. Second-quarter revenue rose 17% organically, existing customers spent more, and management raised full-year guidance. But before treating this as a fresh chance to get in, the acceleration deserves closer reading, because part of it does not actually exist, the company itself expects growth to slow next quarter, and the doubled price now asks the AI story to keep getting better. This is a genuine improvement, and it is no longer priced like a turnaround.

The re-acceleration now has evidence behind it

For several quarters, Twilio's pitch was that artificial intelligence would bend its messaging-and-voice plumbing back toward growth. The second quarter, reported in early August, is the closest thing to proof yet. Revenue hit $1.50 billion, up 22% on a reported basis and 17% organically, a roughly five-point beat over the $1.43 billion analysts had expected, with adjusted EPS of $1.47 clearing the $1.32 consensus. Net revenue retention — a measure of what existing customers spend versus a year earlier — rose to 116% from 114% the prior quarter, the first sign that the multi-year exodus of shared-revenue accounts has begun to reverse. Voice grew more than 20%, software add-ons more than 25%, and Verify more than 30%, with self-serve voice up more than 50% as AI-native builders moved small tests into production. Free cash flow hit a record $353 million for the quarter, and non-GAAP operating margin reached 19%.

That is a real operating improvement, not a product story with no billings behind it. Management used it to raise full-year guidance, lifting organic revenue growth to 13%–13.5% from 9.5%–10.5% and reported growth to 18%–18.5%.

Two qualifiers shrink what that 22% really means

The first qualifier is the gap between the headline 22% and the 17% organic figure. The difference is almost entirely incremental U.S. carrier fees that Twilio passes through to customers — roughly $250 million this year. That money shows up in reported revenue and inflates reported growth, but it is a pass-through: it adds no gross profit, no operating income, and no cash flow. In fact, because it is booked at a low margin, it is dragging reported gross margin down 160 basis points year over year; strip the fees out and gross margin actually improved about 60 basis points. So the flashier 22% number overstates the economics of the underlying business.

The second qualifier looks forward and matters more. Even as the company printed 17% organic growth in the second quarter, its guidance for the third quarter calls for organic growth of just 11%–12%. In other words, management itself sees the current pace as the high point, not the new run-rate, partly because voice and software add-ons face harder year-ago comparisons in the second half. The market is being asked to pay for acceleration at the very moment the company forecasts deceleration.

The multiple has moved ahead of the proof

This is where the stock's move and the business part ways. Twilio trades at about 6.3 times trailing sales against a market value near $37 billion, with free cash flow trailing at roughly $990 million and a healthy net cash position. Put another way, it sells for roughly 30 times this year's guided free cash flow of about $1.14 billion. A year ago, investors could reasonably call the stock cheap for a platform with improving margins; the roughly 125% one-year climb and the 68% year-to-date gain have retired that argument. Today the price assumes the AI cohort keeps expanding and that the cash-flow engine keeps compounding — reasonable hopes, but expectations, not bargains.

The AI evidence is real but still concentrated and small. Twilio still does not break out AI revenue, and the AI-native customers that anchor the pitch — a $9 million account whose software add-ons grew 100% year over year, a $6 million run-rate customer up 65% — are a fast-growing slice of a much larger base, not the base itself. The company can point to a seven-figure agentic deal with Car Finance 247, whose assistant handled roughly 300,000 conversations, as a template. What it cannot yet point to is a number proving AI has become a primary, repeatable driver of the whole platform. The direction is favorable; the scale is still an inference.

The second half is the proof window

None of this makes Twilio a broken story, and none of it argues for panic. The growth re-acceleration is real, free cash flow is strong and still growing, and a roughly $1.7 billion net cash cushion means the balance sheet is not the risk. The honest tension is between a genuinely improved business and a price that has already absorbed much of the improvement. For a holder, the current multiple is reasonable support for a disciplined hold, but little of the estimate-beating upside is still free. For a watcher, the case for paying up improves only if the second-half comparisons bend toward the 17% second-quarter print rather than the 11%–12% that management guided — and if the AI-native cohort, still a small minority of revenue, keeps outgrowing everything else. If organic growth holds nearer the high end of that range through the year and margins keep climbing, the premium will have earned itself. If growth slips back toward the low teens while the stock stays at 6 times sales, the market will have paid for a best-case quarter in advance.

That is the way to read this rally: a real operating improvement that the market has largely priced. The opportunity that existed at $99 a year ago is not the opportunity at $240. The next two quarters, not the last one, decide whether this is the resumption of a compounder or the top of a good run.