Investors shift focus away from generic AI SaaS tools

Investors have poured billions of dollars into artificial intelligence companies over the past few years, as the technology continues to hold sway in the Valley and thus the world. However, venture capitalists (VCs) are shifting their attention away from “thin” software-as-a-service (SaaS) products. According to Aaron Holiday, a managing partner at venture capital firm 645 Ventures, investors are basically no longer interested in startups building thin workflow layers, generic horizontal tools, light product management, and surface-level analytics. Abdul Abdirahman, an investor at venture capital firm F-Prime, added that generic vertical software without proprietary data moats is also no longer popular.

This shift reflects a broader decline in the value of thin AI wrappers, generic productivity tools, and basic CRM clones. Igor Ryabenkiy, founder and managing partner at venture capital firm AltaIR Capital, noted that the barrier to entry has dropped, making it harder to build a real moat. Ryabenkiy explained that if a startup’s differentiation lives mostly in its user interface and automation, that is no longer enough.

The rise of agents is also changing how investors evaluate software stickiness. Jake Saper, a general partner at venture capital firm Emergence Capital, pointed to the difference between developer tools like Cursor and Anthropic’s Claude Code as an early indicator of this shift. Saper noted that while one tool owns the developer’s workflow, the other simply executes the task, with developers increasingly choosing execution over process. Saper explained that while getting humans to work inside a software platform used to be a strong defense, the rise of agents changes the dynamic. “Pre-Claude, getting humans to do their jobs inside your software was a powerful moat, but if agents are doing the work, who cares about human workflow?” Saper said.

Saper also noted that integrations are increasingly becoming less popular. This shift is driven by Anthropic’s model context protocol (MCP)—a standard for connecting AI models to external data—which makes it easier to connect models to external systems without building custom integrations.

To remain attractive, startups must demonstrate deep workflow integration and move toward flexible pricing. Ryabenkiy stated that rigid per-seat pricing models will be harder to defend, while consumption-based models make more sense.

Why it matters

Investors are reallocating capital toward businesses that own workflows, data, and domain expertise, while moving away from products that can be easily replicated by AI-native teams.