Daisy Cai of B Capital said AI agents represent a "restacking" of software rather than an existential threat, implying the industry is shifting rather than deteriorating. She noted that traditional SaaS models are being challenged as enterprise buyers prioritize outcomes and efficiency over simply adding seats. The commentary is directional rather than event-driven, with limited immediate market impact.
The key second-order effect is not that software disappears, but that value migrates from user-count pricing to workflow ownership and model orchestration. That tends to compress value for horizontal SaaS with bloated seat-based ARR while expanding it for vendors that sit closest to data, permissions, and execution layers; in practice, the market will reward products that can prove measurable labor substitution within one budget cycle. This creates a bifurcation: “workflow winners” should sustain premium multiples, while point solutions and low-switching-cost apps face slower net retention and higher churn risk.
The near-term losers are likely vendors whose pricing power depends on headcount growth rather than output metrics. If buyers can deploy agents to automate 10-20% of routine tasks, CFOs will demand payback periods under 12 months, which forces a reset in contract structure and can pressure gross retention before revenue growth visibly rolls over. The second-order beneficiary is infrastructure: cloud, vector databases, observability, and identity/security layers should see increased spend even if app-layer budgets get rationalized, because agentic systems increase tool calls, inference usage, and governance requirements.
The contrarian miss is that “restacking” may be slower than the narrative implies because enterprise adoption is constrained by reliability, auditability, and integration debt. That means the market could over-discount SaaS impairment in the next 6-12 months while underestimating the monetization power of incumbents that own distribution and compliance. The biggest risk to the thesis is a product-cycle breakthrough that turns agents from assistive tools into durable replacements; until then, this is more a margin and pricing mix issue than a full secular disruption.
From a portfolio perspective, the best setup is to fade the most expensive seat-count SaaS names into strength and own the picks-and-shovels of agent deployment. I would expect the first real earnings impact to show up over the next 2-4 quarters in slower billings growth and softer expansion revenue, not headline revenue collapse. That creates a favorable window for relative-value positioning before consensus fully adjusts its model assumptions.
AI-powered research, real-time alerts, and portfolio analytics for institutional investors.
Request TrialOverall Sentiment
neutral
Sentiment Score
0.10
