Investors are posting what they are no longer looking for in AI SaaS companies

Investors have poured billions into AI companies over the past few years, as the technology continues to take over the Valley and thus the world. But not all AI companies attract investor interest.

In fact, even though every company these days is rebranding to include “artificial intelligence” in its name, some startup ideas are no longer favored by investors. TechCrunch spoke with VCs to find out what investors aren’t looking for in startups that offer AI software as a service anymore.

Popular SaaS categories for investors now include startups that are building native AI infrastructure, SaaS verticals with proprietary data, business systems (those that help users complete tasks), and platforms deeply integrated into mission-critical workflows, according to Aaron Holliday, managing partner at 645 Ventures.

But he also provided a list of companies that are pretty boring for investors these days: startups that build thin workflow layers, generic horizontal tools, light product management, surface-level analytics — basically, anything an AI agent can do right now.

Public vertical software “without private data trenches” is no longer common, F Prime investor Abdul Rahman added, and Igor Ryabenki, founder and managing partner at AltaIR Capital, delved into this point. Investors aren’t interested in anything that, in fact, doesn’t have a lot of product depth, he said.

“If your differentiation lives mostly in the user interface [user interface] “And automation is no longer enough. The barrier to entry has lowered, making it more difficult to build a real trench.”

New companies entering the market now need to build on “real ownership of the workflow and a clear understanding of the problem from day one,” he said. “Large code bases are no longer an advantage. What matters most is speed, focus, and the ability to adapt quickly. Pricing also needs to be flexible: strict per-seat models will be difficult to defend, while consumption-based models make more sense in this environment.”

Jake Saper, general partner at Emergence Capital, also had thoughts on ownership. For him, the differences between Corsor and Claude Code are “the canary in the coal mine.”

“One has the developer workflow, and the other just does the job,” Saper continued. “Developers are increasingly choosing implementation rather than process.”

Any product that deals with “workflow stickiness” — that is, trying to get as many human customers as possible to continuously use the product — may find itself in an uphill battle when agents take over the workflow, he said.

“Before Claude, having humans do their jobs inside your software was a powerful moat, but if agents are doing the work, who cares about the human workflow?” He told TechCrunch.

He also believes integrations are becoming less common, especially as Anthropic’s Model Context Protocol (MCP) makes it easier than ever to connect AI models to external data and systems. This means that someone doesn’t need to download multiple integrations or create their own client integrations; They can only use MCP.

“Being the conductor has been a trench,” Saper said. “Soon, it will be of use.”

And including “workflow automation and task management tools that enable the coordination of human work is no longer less important if, over time, agents just perform tasks,” Abdel-Rahman said, citing examples, particularly public SaaS companies whose stocks have fallen as new AI-based startups emerge with better, more efficient technology.

The SaaS companies that are struggling to raise money right now are the ones that can be easily replicated, Ryabenki said.

“General productivity tools, project management software, core CRM clones, and thin AI wrappers built on top of existing APIs fall into this category,” he said. “If a product is mostly an interface layer with no deep integration, proprietary data, or knowledge of the processes involved, strong AI teams can rebuild it quickly. That’s what makes investors wary.”

Furthermore, what remains attractive about SaaS is the depth and expertise, with built-in tools for critical workflows. He said companies should now look at integrating AI deeply into their products and update their marketing to reflect this.

“Investors are reallocating capital toward companies that have workflow, data, and domain expertise,” Ryabenki said. “Away from products that can be imitated without much effort.”