SaaS is repricing from per-seat to per-outcome. AI agents don't need seats. Traditional SaaS trades at 8-15x revenue because recurring per-seat revenue is predictable. Outcome-based services businesses trade at 2-4x. That multiple compression is the selloff. The disruption is being forced by AI agent startups. They're small, early, and have zero legacy revenue to protect. They're pricing on outcomes from day one because they have no seat-based revenue to cannibalize. But the public SaaS companies with real moats (proprietary data, embedded workflows, and high switching costs from years of customer data locked in their systems) can choose to disrupt themselves first. If they shift to outcome-based pricing, they're not shrinking. They're going from selling software seats to eating into the massive global services market, which is multiples larger than their current TAM. The catch: it requires tanking your near-term stock to win the long game. Most management teams and boards won't stomach that. The ones that do will own a much bigger pie. The ones that don't will get eaten by startups that never had a pie to protect.