What Investors No Longer Want to See in AI SaaS Startups

Investors have poured billions of dollars into AI companies over the last several years, as the technology continues to dominate Silicon Valley and, by extension, much of the broader business world. But that does not mean every AI company is capturing investor enthusiasm.

In fact, even though it can seem as if nearly every company is now rushing to add “AI” to its branding, certain startup concepts have clearly fallen out of favour with investors. We spoke with venture capitalists to better understand which types of AI software-as-a-service startups no longer generate the same level of excitement.

Among the SaaS areas that investors still find compelling are startups building AI-native infrastructure, vertical SaaS companies with proprietary data, systems of action that help users actually complete work, and platforms deeply integrated into mission-critical workflows, according to Aaron Holiday, a managing partner at 645 Ventures.

At the same time, he pointed to a group of startups that investors now view as relatively uninteresting: companies building thin workflow layers, generic horizontal tools, lightweight product management software, and surface-level analytics products—essentially, products that an AI agent can now handle on its own.

Abdul Abdirahman, an investor at F-Prime, added that generic vertical software “without proprietary data moats” is no longer especially attractive, and Igor Ryabenkiy, founder and managing partner at AltaIR Capital, expanded on that idea. He said investors are increasingly uninterested in businesses that lack real product depth.

“If your differentiation lives mostly in UI [user interface] and automation, that’s no longer enough,” he said. “The barrier to entry has dropped, which makes building a real moat much harder.”

According to him, new startups entering the market now need to be built around “real workflow ownership and a clear understanding of the problem from day one.” He added that “massive codebases are no longer an advantage. What matters more is speed, focus, and the ability to adapt quickly. Pricing also needs to be flexible: rigid per-seat models will be harder to defend, while consumption-based models make more sense in this environment.”

Jake Saper, a general partner at Emergence Capital, also pointed to the importance of ownership. In his view, the difference between Cursor and Claude Code is the “canary in the coal mine.”

“One owns the developer’s workflow, the other just executes the task,” Saper said. “Developers are increasingly choosing the execution over process.”

He said any product built around “workflow stickiness” — software designed to attract large numbers of human users and keep them continuously active within the product — could face a difficult road ahead as AI agents begin to take over more of the workflow itself.

“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?” he said.

He also argued that integrations are becoming less compelling, especially as Anthropic’s model context protocol (MCP) makes it easier to connect AI models to external data sources and systems. In practice, that means users may no longer need to download multiple integrations or spend time building their own customer integrations; instead, they can rely on MCP.

“Being the connector used to be a moat,” Saper said. “Soon, it’ll be a utility.”

Abdirahman said that “workflow automation and task management tools that enable the coordination of human work become less necessary if, over time, agents just execute the tasks,” pointing to public SaaS companies in particular, whose share prices have fallen as AI-native startups emerge withfasterr, more efficient technology.

Ryabenkiy said the SaaS businesses having the hardest time raising capital at the moment are the ones that can be copied too easily.

“Generic productivity tools, project management software, basic CRM clones, and thin AI wrappers built on top of existing APIs fall into this category,” he said. “If the product is mostly an interface layer without deep integration, proprietary data, or embedded process knowledge, strong AI-native teams can rebuild it quickly. That is what makes investors cautious.”

Overall, he said the parts of SaaS that remain appealing are depth, expertise, and tools that sit inside essential workflows. Ryabenkiy added that companies should now look at integrating AI more deeply into their products and update their marketing to reflect that shift.

“Investors are reallocating capital toward businesses that own workflows, data, and domain expertise,” Ryabenkiy said. “And away from products that can be copied without much effort.”