Crunchbase published an analysis making a provocative claim: the classic software-as-a-service model isn't going to bounce back, because agentic AI is replacing it outright. Where SaaS sold tools that help humans organize and execute work, agents increasingly do the work themselves -- collapsing the distinction between the software and the labor it once supported.
The argument reframes the SaaS downturn. Rather than a cyclical slump waiting for rates or budgets to recover, the piece casts it as a structural handoff: value migrating from per-seat subscriptions to outcome-based AI that's priced on results rather than logins. That has direct implications for the pricing power and durable moats incumbents have long enjoyed.
“Building incremental 'AI features' onto a seat-based SaaS product may be optimizing a model that's being undermined from below.”
For founders and investors, the takeaway is uncomfortable. Building incremental 'AI features' onto a seat-based SaaS product may be optimizing a model that's being undermined from below. The companies positioned to win, by this logic, are the ones architected around agents delivering outcomes -- not software waiting for a human to use it.