Table of Contents
Last year, Bandwidth earned $20 million from political texting. Could it match its election-year driven 2024 growth? Check. Would it explain the dip in political texting revenue in a non-election year? Check. Could its voice business deliver profitable, sustainable growth? Big check. In one earnings call, Bandwidth answered my three big questions.
Bandwidth’s revenue hit $180M (+9%), EBITDA reached $22M (+17%), and AI-enhanced voice services gained solid traction. Maestro’s AI-driven orchestration is now directly fueling voice revenue growth, validating Bandwidth’s strategic bet on AI. Messaging grew a disciplined 7% by emphasizing profitability over growth. It even raised guidance for the rest of the year.
Well done.
The Good
Bandwidth exceeded guidance in Q2 and showed confidence and ambition by increasing 2H revenue growth guidance to 10% (Q3) and 13% (Q4).
Enterprise voice revenue surged 29%, directly tied to robust AI voice adoption and major strategic wins. Notably, every enterprise deal this quarter involved Maestro, validating Bandwidth’s AI orchestration platform as a growth engine and ARPU driver.
Each AI-enabled call triggers multiple simultaneous service streams (e.g., transcription, sentiment analysis, fraud detection), generating 3x–4x per-call revenue. Customer retention remains exceptional (>99%), net retention hit 112%, and average annual revenue per customer rose to a record $230K.
The Global Voice Plan (GVP) segment posted its strongest growth in years (7% vs. 2% previously), propelled by increased AI-driven use cases in UCaaS and CCaaS.
The recent NiCE/Cognigy merger is a win for Bandwidth. Both were existing partners, and now their combined user base can seamlessly leverage Bandwidth’s AIBridge via Maestro. This is exactly how a strategic partnership should work: tighter integrations, shared customers, and clear revenue upside.
In messaging, Bandwidth focused on profitable, high-margin formats—explicitly mentioning RCS and RBM—and prioritizing larger enterprise customers demanding reliability, compliance, and scalability.
The Interesting
Adoption of AI services on the Maestro platform is progressing gradually but predictably. Customers typically start with initial cloud migrations (using Bandwidth’s Mist platform) and incrementally layer advanced AI features over time, providing a clear road map for continued incremental revenue growth.
Recent competitor price increases have created strategic customer acquisition opportunities for Bandwidth. Bandwidth is actively capitalizing on this shift but emphasized a disciplined approach to preserve profitability.
The Unknown
Bandwidth’s valuation remains puzzlingly disconnected from recent market benchmarks, specifically the nearly $1 billion NiCE paid for Cognigy. Despite Bandwidth’s far larger revenue footprint (~$750M annually), its market valuation is roughly half of Cognigy’s. While none of this should be considered stock advice, the stock continues to trade at a discount to revenue.
The company acknowledged this discrepancy, attributing it to the market’s incomplete recognition of its strategic global infrastructure role for AI voice. How quickly the market bridges this valuation gap remains uncertain.
The brevity of remarks on RCS and RBM suggests these formats aren’t yet driving meaningful messaging revenue. Still, Bandwidth’s 7% messaging growth outpaces its cohort. But, if competitor pricing is creating growth opportunities, does that signal messaging commoditization—or is the payoff from product investments still pending?
Finally
In BlackBerry, there’s a scene where RIM CEO Jim Balsillie explains to Mike Lazaridis how the AT&T and iPhone partnership transforms phones from “just voice” into “selling data.” In the movie, like in real life, it marks the beginning of the end for the phone maker.
For the archipelago of companies handling voice between consumers and businesses, AI is creating an iphone like moment. It is making voice a data-like revenue multiplier.
For Bandwidth, voice is no longer just voice. One minute of conversation now begets one minute of transcription, one minute of sentiment analysis, and one minute of real-time response. AI has once again turned voice into both the medium and the app.