Jefferies upgraded Twilio from Hold to Buy on Monday, raising its price target from $125 to $160, as the firm argued the communications platform is poised to become a central piece of infrastructure in the emerging voice AI economy. Analyst Samad Samana, who issued the note alongside Jeremy Sahler, wrote that the firm has “greater conviction in the role TWLO will play in the Voice AI tech stack,” adding that “even a modest increase in traction can serve as a sustainable tailwind for gross profit dollar growth for years to come and drive upside to consensus.” The $160 target implies roughly 21 times Jefferies' 2027 free cash flow estimate.
The Voice AI Thesis | From 3 Cents to 21 Cents Per Call
At the heart of the upgrade is a bet on how voice AI will reshape Twilio's economics. In traditional voice calls, Twilio monetizes at the lowest layer of the telecom stack, netting roughly three cents per call. But with products like ConversationRelay, which bundles speech-to-text, text-to-speech, and orchestration tools, the firm estimates Twilio could capture about 21 cents per three-minute voice AI call, a more than sevenfold increase in revenue per interaction with materially higher margins.
| Metric | Detail |
|---|---|
Previous Rating | Hold, $125 price target |
New Rating | Buy, $160 price target |
Implied Valuation | ~21x 2027 free cash flow estimate |
Traditional Voice Revenue | ~$0.03 per call |
Voice AI Revenue (ConversationRelay) | ~$0.21 per 3-minute call |
Revenue Uplift | ~7x per interaction |
Projected Gross Margin Improvement | ~115 basis points by 2028 |
Jefferies sees voice AI as “the convergence of cloud telephony, real-time media streaming, speech-to-text, LLM inference, and text-to-speech to enable fully automated, conversational call interactions,” with monetization concentrated at the orchestration layer where Twilio operates. The analysts project the shift could drive approximately 115 basis points of gross margin improvement by 2028.
ConversationRelay | How Twilio Moves Up the Stack
ConversationRelay is the product at the center of the thesis. Rather than simply routing a phone call from one endpoint to another, the product sits between the caller and an AI agent, handling the real-time conversion of speech to text, passing that text to a large language model for processing, and converting the LLM's response back to speech. Twilio manages the latency-sensitive orchestration, the telephony connection, and the handoff between each component in the pipeline.
This positions Twilio not as a commodity carrier but as the middleware layer that makes voice AI work at scale. The company does not need to build its own LLM or speech model. It profits by being the infrastructure that connects those components to the existing telephone network, which remains the primary way businesses interact with customers for support, sales, and scheduling.
The economic logic is straightforward. A traditional Twilio voice call generates revenue from a per-minute telephony charge. A ConversationRelay call generates revenue from the telephony charge plus orchestration fees, speech-to-text processing, text-to-speech processing, and API metering for each round-trip between the caller and the AI agent. Every layer in the stack that Twilio touches adds incremental revenue to the same underlying call.
Market Context | Voice AI Adoption Is Still Early
Jefferies acknowledges that voice AI adoption is in its early stages, which is precisely why the analysts framed the thesis around modest traction rather than aggressive penetration assumptions. Enterprise call centers, healthcare appointment systems, and financial services automated lines are among the earliest adopters, but the total addressable market extends to any business that handles inbound or outbound phone calls, a category that includes millions of companies worldwide.
Competitors exist at every layer. Google Cloud, Amazon Web Services, and Microsoft Azure all offer speech-to-text and text-to-speech APIs. Startups like Bland AI, Retell, and Vapi are building voice AI platforms that compete directly with ConversationRelay. But Twilio's advantage, according to Jefferies, is its existing installed base. The company already handles billions of voice and messaging interactions per year for customers ranging from Uber to Morgan Stanley. Layering voice AI onto those existing connections is a distribution advantage that pure-play startups cannot replicate.
What Could Go Wrong | Risks to the Bull Case
The upgrade is not without caveats. Jefferies noted that Twilio's core communications business has been growing in the low single digits, and the company faces ongoing pressure to demonstrate that its Segment customer data platform acquisition was worth the $3.2 billion it paid in 2020. If voice AI adoption stalls or enterprise customers build their own orchestration layers using open-source tools, the margin expansion thesis could take significantly longer to materialize.
There is also pricing risk. As more providers enter the voice AI orchestration market, per-call economics could compress before Twilio reaches the scale needed to sustain the projected margin improvement. And regulatory scrutiny of AI-powered phone calls, particularly around disclosure requirements and consumer consent, could slow adoption in industries like healthcare and financial services where voice interactions are subject to compliance frameworks.
For now, though, Jefferies' thesis is that Twilio's position in the voice AI stack is underappreciated by the market. At 21 times forward free cash flow, the stock prices in modest growth. If ConversationRelay gains traction, the upside is that Twilio transforms from a low-margin telecom utility into a high-margin AI infrastructure provider, and the market has not yet priced in that possibility.
Filed under
Discussion
Every comment appears live in our Discord server.
Join to see the full conversation and connect with the community.
Comments sync to our ObjectWire Discord · Jefferies Upgrades Twilio to Buy, Betting on Voice AI Growth.
Written by
Alfanasa