Taking stock of the SaaSpocalypse - Foundation Capital
This month's software selloff isn't about cyclical slowdown—it's about terminal value: whether SaaS business models built on seats and interfaces remain valid when agents do the work and context graphs replace data gravity as the moat.
Read OriginalMy Notes (2)
SaaS lock-in has historically been built on two kinds of friction: the cost and risk of migrating data at scale, and the user habituation that comes from teams building workflows around a specific interface. Agents weaken the second kind: users interact with the agent rather than the underlying system directly, which makes the interface less sticky. This dynamic affects SaaS broadly: any tool whose value lived primarily in its interface is vulnerable.
What replaces friction as the moat is context: the organizational decision histories and reasoning patterns that accumulate as an AI system learns how a specific company thinks and operates.
when a foundational cost collapses (compute, bandwidth, distribution), the economy tends to expand through the savings rather than shrink around them. The incumbents who relied on the old economics are hit hard. But the total market grows, and the productivity gains flow into new markets, new products, and new jobs that didn’t exist before
Summary used for search
TLDR
• The real threat isn't that software dies, but that value migrates from systems of record to the agent layer above them—turning incumbents into commodity infrastructure
• Incumbent systems store outcomes; agents need "context graphs" that capture organizational reasoning patterns and decision histories—a fundamentally different architecture
• Mid-tier SaaS ($100M-$1B ARR) faces existential pressure: too big to pivot fast, too small to absorb repricing while investing in transformation
• AI products run 25% gross margins vs 70-90% for traditional SaaS due to inference costs, while category boundaries collapse as agents handle end-to-end workflows
• The market will be 10x bigger in 10 years, but AI-native startups will capture most growth because they're unburdened by legacy architecture and revenue bases to protect
In Detail
Foundation Capital argues that February 2025's software selloff represents a fundamental repricing around terminal value, not cyclical headwinds. Unlike 2022's interest rate-driven correction, investors are now questioning whether SaaS business models remain valid as AI matures. The core thesis: the software market will expand 10x over the next decade, but AI-native startups—not incumbents—will capture most of that growth.
The piece identifies five structural pressures breaking traditional SaaS: (1) The seat model collapses as teams do more with fewer people and pricing shifts to usage/outcomes; (2) Margins compress from 70-90% to ~25% due to inference costs; (3) Category boundaries dissolve as agents handle end-to-end workflows that previously required separate point solutions; (4) Value migrates to the agent layer, turning systems of record into commodity infrastructure; (5) The nature of moats shifts from interface stickiness and data gravity to "context graphs"—reasoning layers that capture how companies think, not just what they've done.
The author's key insight: incumbent systems were optimized for human workforces and store outcomes, but agents need decision histories and reasoning patterns. Retrofitting this "context graph" capability onto legacy architecture is fundamentally harder than building from scratch. Mid-tier SaaS ($100M-$1B ARR) faces the most pressure—too big to move fast, too small to absorb repricing. The winners will be AI-native startups tackling massive problems that were previously too complex or expensive, and incumbents who successfully acquire their way into the new paradigm. The cost of building has collapsed, making incremental ideas insufficient when two people and a compute budget can now do what used to require hundreds of engineers.