Notes on where seat-based pricing is going (paid.ai)

The note argues that traditional seat-based SaaS pricing is not just under threat — it’s becoming irrelevant as AI agents and automation replace human seat-based work. For two decades ARR scaled with headcount (more employees → more seats → more predictable revenue), but three structural shifts broke that model: work executed by APIs/agents (which don’t occupy seats), much smaller teams achieving far more via automation, and built‑in AI features that reduce human usage minutes. The economic value that used to live in licenses has migrated to compute and agent execution, so seat counts and login metrics now measure inertia, not value. Practically, vendors are already moving to “per-work” pricing with three emerging archetypes: usage-based (API calls, tokens, compute minutes tied to workload volume), outcome-based (billing by business results like leads verified or tickets resolved), and agent-based (charging per autonomous agent or synthetic labor unit). Expect a messy transitional period — hybrid seat+usage models, credits and caps — similar to the early days of cloud billing, before convergence toward work-per-unit-time and outcome billing. The implication for product, telemetry and GTM teams is profound: rethink signals, instrument work outputs not logins, and realign pricing to what software actually delivers (work/outcomes) rather than who touches it.