Sinch's Voice AI Bet Meets Its Divestment Door
Sinch's Voice AI Bet Meets Its Divestment Door
Sinch reported a quarter that looked, on the surface, like more of the same: 5% organic revenue growth, 4% organic gross profit growth, an Americas engine growing 7% in revenue and 10% in gross profit, and an all-time-high EBITDA margin of 12.9%. But under the hood, three things changed: the growth narrative moved decisively from RCS toward voice AI, management opened the door to divestments, and the buyback that was supposed to be the obvious use of cash remained paused. The result is a company trying to convince investors it has a second act before its first act's margin tailwind runs out.
The Quarter That Wasn't Quite the Guidance
The headline numbers were fine, not spectacular. Americas is two-thirds of the business and did the heavy lifting again, with messaging and e-mail APIs driving growth and network connectivity benefiting from the ongoing shift from legacy TDM to IP technology. EMEA remained the drag, down 3% in revenue and 2% in gross profit, while APAC grew revenue 5% but saw gross profit decline on margin compression in Australia and a tough comparison in India.
Management leaned hard on the idea that underlying profitability is stronger than the reported gross margin suggests. Certain U.S. suppliers raised costs, and Sinch passed those through to customers at cost — adding roughly 2 percentage points to organic revenue growth but diluting gross margin because the traffic carries no margin. Sofia Ohlander framed the adjustment clearly: “In Q3, we have an all-time high EBITDA margin since we were listed at 12.9%.” — Sofia Ohlander, Senior Vice President, Corporate Control · 2026-11-05 Cash flow was genuinely strong at SEK 787 million, rolling 12-month cash conversion was 75%, and net debt to adjusted EBITDA fell to 1.7x. The balance sheet is no longer the problem.
The problem is the growth algorithm. Sinch has a midterm target of 7% to 9% organic growth by 2027, and it is running at 4% organic gross profit growth. The biggest current margin tailwind — the TDM to IP transition in network connectivity — is nearly complete. Management expects another couple of percentage points of gross margin expansion over the next two to three quarters, then the tailwind fades. Asked what replaces it, Jonas Dahlberg pointed to revenue growth and the higher-margin e-mail mix. That is plausible, but it means the next leg of the story must come from volume, not from another self-help margin lever.
Voice AI Is the New Headline — and It's Still Small
The keyword rotation in this quarter is striking. Sinch's top three curated themes are now voice AI, voice API, and voice network. That is a different company from the one that spent the past two years talking up RCS and conversational messaging as the next generation of SMS.
Dahlberg made the case that voice is back as a channel, and that Sinch owns the infrastructure to monetize it: the largest independent carrier-grade voice network in the U.S., more than 250 billion voice minutes a year, direct connections to all relevant U.S. carriers, and a pool of 160 million phone numbers. The new voice API version 2 is aimed squarely at agentic AI developers. “So if you combine our voice API with our voice network infrastructure, we have what I would be confident to say the best home for agentic voice in the market.” — Johnas Dahlberg, Acting CEO and CFO · 2026-11-05
Sinch is also pushing into agentic development tools, making its APIs available inside Anthropic's Claude, OpenAI's Codex, Google Agentspace, Lovable, Eleven Labs, and — most notably — SpaceX Super Intelligence's Grok Build. That last one lands right on top of a global theme: Grok Bot is one of the market's hottest curated keywords this quarter. Sinch is trying to position itself as the communications plumbing behind that wave, not as a bystander.
But the monetization is still early. When asked when AI would meaningfully contribute, Dahlberg did not overreach: “We expect it to meaningfully contribute to growth next year. So maybe low single-digit contributions to growth next year.” — Johnas Dahlberg, Acting CEO and CFO · 2026-11-05 That is honest, but it is also a reminder that this is not yet a revenue story. It is a narrative option. Prior calls had the same shape: a year ago, Laurinda Pang said of AI, “We're seeing a little bit of that at this point, but we do think that there is opportunity in the future.” — Laurinda Y. Pang, CEO · 2025-07-22 The emphasis is new; the timing keeps slipping to the right.
RCS Cooled, Fixed Fees Still Drag
The quiet admission in the quarter was on RCS. Sinch now has full Tier 1 RCS coverage in North America, 84 direct connects globally, and traffic that doubled year-over-year to 1.4 billion messages. North American traffic is up fivefold. Those are good operating metrics. But management conceded the commercial ramp is slower than expected.
We have to be honest, this moves slower than we expected. And a lot of this is onboarding friction with operators. ... It's still low single-digit percentage of the total messaging mix.
That is a meaningful shift from the prior stance. In November 2025, Pang was still framing RCS as progress: “We are seeing good progress in RCS. We're seeing great volume increases in RCS, but it's actually not impacting the financials just yet.” — Laurinda Y. Pang, CEO · 2025-07-22 A year later, the volume is larger but the financial impact is still not there, and the language has cooled from “next generation of SMS” to “slower than we expected.” RCS has not disappeared from the story, but it is no longer the headline.
The old drag remains too. Fixed fee operator contracts are still a top-five keyword and still weigh on EMEA comparisons, though they are now a low-single-digit share of revenue and gross profit and are stabilizing sequentially. EMEA has a new leader and a growing pipeline, but the turnaround is still a few quarters away in the numbers.
The Capital Allocation Pivot
The most consequential new language was not about AI. It was about divestments. Dahlberg said Sinch is open to selling adjacent businesses that came with past acquisitions but do not fit the go-forward portfolio. That is a new communication, and it matters because Sinch has spent years defending its roll-up history.
It's not the entire product portfolio. What this is about is — you know that Sinch made several major acquisitions a couple of years back. With those acquisitions also came, let's say, adjacent businesses, which when we combine this into one Sinch and would position ourselves for an AI future, may not — it's good businesses, but they may not really fit the portfolio. So they could find a better home maybe somewhere else.
At the same time, the buyback is paused. Sinch repurchased 15% of its shares last year and has a mandate for up to 10% more, but it made no repurchases this quarter. Management says the Board is optimizing across buybacks, deleveraging, and M&A. That is a different tone from prior calls, when Dahlberg said, “We're ready for M&A. And so we're looking at the market. And if the right opportunity comes up, we will act on that to drive consolidation.” — Jonas Dahlberg, Acting CEO and CFO · 2026-07-22 The balance sheet is strong enough to do any of the three. The question is which story the Board wants to tell.
There is also a governance overhang. Dahlberg is acting CEO and CFO, and the permanent CEO search has dragged on. Asked for an update, he deflected with a joke about giving the analyst his chairman's phone number. That is not a fundamental data point, but it reinforces the sense of a company in transition.
What Actually Changed
Sinch is not a broken business. It has a two-thirds Americas franchise, a high-margin e-mail product, a valuable voice network, strong cash flow, and a leverage ratio that is now comfortably inside target. The quarter proved the company can generate cash and defend margins.
But the growth narrative is being rewritten in real time. API Platform messaging and e-mail still pay the bills. Voice AI is the new hope. RCS is no longer the near-term catalyst. Divestments are now on the table. And the buyback is no longer the default use of capital. For investors, the market will have to underwrite a simple question: is agentic voice a real second act for Sinch, or is it a bridge until the next margin tailwind arrives? The answer will not come from this quarter's income statement. It will come from whether those low-single-digit AI contributions actually show up next year.