S15.9 · Telecom & Connectivity
Cloud voice, messaging and contact-center platforms where agentic AI automates the per-seat revenue unit itself, forcing a shift to consumption pricing.
Unified communications and cloud voice providers sell cloud-based voice, messaging and contact-center platforms to enterprises. The global UCaaS market is estimated at $68.42B in 2025 (Cognitive Market Research); contact-center-as-a-service (CCaaS) is smaller and historically faster-growing, commonly estimated in the $10-15B range, though with wide variance across research firms. Every other segment in this sector experiences AI as a repricing of demand for connectivity. This one does not. Here, AI automates the product itself — and the revenue unit with it.
UCaaS grows at a high-single to low-double-digit rate. CCaaS has historically outgrown it, but AI-agent substitution for human contact-center staff now sits as a headwind against seat-based revenue forecasts. Demand concentrates in North America and Europe; vendor research and development is similarly US-centered. Concentration is moderate: a handful of scaled platforms — Microsoft Teams, RingCentral, Zoom, 8x8, Genesys, NICE and Five9 — above a fragmented small-business long tail.
Upstream, the segment rides on underlying connectivity and cloud infrastructure; downstream, buyers are enterprise IT and customer-experience organizations. This is the most software-like business in the sector — recurring per-seat pricing, 60-70% gross margins, low capital intensity, and no spectrum or facilities constraint gating entry the way it does elsewhere in telecom. That missing structural moat is precisely the exposure: nothing stops a well-capitalized software entrant from building a competing platform, which is exactly what AI-native challengers are now doing.
The mechanics run straight through the pricing unit. Agentic AI directly automates the contact-center agent seat — the CCaaS revenue unit itself. As AI absorbs a rising share of tier-1 contact volume, it compresses the addressable seat count that CCaaS revenue has historically been sized on. That is a structural challenge to the segment's own pricing model, not merely an efficiency gain for the customers buying the service; the dollars saved on seats do not automatically come back to the vendor in another form.
The vendor responses split cleanly. Repricing from per-seat to per-resolution and consumption-based models is the defensive move — an attempt to capture the value of automation before customers displace the seat relationship entirely. AI-native customer-experience platforms entering the market to sell orchestration rather than seats are the expansionary counter-move. The shift runs deep enough that platforms are repositioning their core offering as AI agent orchestration rather than a seat-licensed communications tool — a genuinely new category boundary, and one already underway rather than a future possibility.