Key insights
- Bandwidth Inc. (BAND), a CPaaS company, projects strong growth in revenue, margins, and EPS through 2026. Key drivers include high revenue retention, expansion in enterprise voice, and the adoption of its AI orchestration platform, Maestro. A share repurchase program adds further support. This positive outlook could moderately influence investor sentiment in the cloud communications sector.
bandwidth inc (BAND) is a CPaaS company — cloud communications infrastructure. they own and operate their own global network across 65+ countries covering 90% of global GDP. unlike twilio or vonage who rent third-party networks, bandwidth owns the pipes.
what they actually do
enterprises and platforms use bandwidth's APIs to embed voice, messaging, and emergency services into their software. think of them as the carrier layer that microsoft, google, zoom, ringcentral, five9, and genesys all sit on top of.
three segments: global voice plans (UCaaS/CCaaS platforms), enterprise voice (direct to Global 2000), and programmable messaging (A2P SMS/MMS).
the numbers — FY2025
revenue: $561M total — $414M voice (~60%), $147M messaging (~40%)
market cap: ~$1.97B at $61.39
gross margin trajectory: targeting 60%+ by end of 2026
incremental gross profit yield: 82% in 2025 — every new revenue dollar is highly profitable
non-GAAP EPS guidance 2026: $1.66–$1.74 (19% growth)
EBITDA growth guidance 2026: ~30% YoY
cumulative FCF target (2023–2026): $125M — already hit by end of 2025
$80M share repurchase authorized Q4 2025
retention and customer quality
98.8% revenue retention (12-month)
100% enterprise customer name retention in 2025 — zero churn
top 20 accounts: median tenure of 12 years
average revenue per customer: $232K vs $171K three years ago
40%+ of 2025 enterprise voice growth came from accounts added in the past 3 years
the AI angle
management launched maestro — an orchestration platform that lets enterprises deploy AI voice agents (from openAI, google, native CCaaS AI, or any third-party) without custom coding. the key insight: every AI call still generates voice minutes on bandwidth's network. AI adoption is a usage driver, not a threat.
real deployments already live: a digital commerce customer running 100% AI voice for food venue ordering nationwide, a top-10 U.S. bank modernizing contact center with AI, and a major insurance group deploying google conversational AI for claims.
software services (maestro, number reputation management, call assure) hit a $15M annualized run rate by year-end vs $10M expectation. now attached to every million-dollar-plus deal.
why the stock is interesting here
management's own framing: "trading near 2017 IPO prices with 3x the revenue, 7 points higher gross margins, and a 65-country footprint."
only 24% of enterprises have moved business-critical communications to the cloud (IDC). 90% plan to by 2026 (gartner). bandwidth is early in a long migration cycle.
no single customer is more than 10% of revenue. the base is diversified across technology platforms, enterprises, and SaaS.
the risks
messaging revenue is cyclical — 2024 and 2026 are election years which boost SMS volume, 2023 and 2025 were softer. revenue CAGR will likely miss the 15–20% target set in 2023 because of this.
top UCaaS customers (microsoft, zoom, google) are also competitors for enterprise direct deals. if any of them vertically integrate their carrier layer, it's a real risk.
hyperscalers (aws, google, microsoft) have unlimited R&D budgets and could out-invest bandwidth in AI voice over time.
the simple thesis
owned infrastructure + enterprise-grade APIs + AI orchestration layer, trading at a discount to where it was at IPO despite being a fundamentally stronger business. the question is whether maestro becomes a real software revenue line or stays a feature.
anyone tracking this or have a view on the TWLO comparison?