AI agents are starting to eat SaaS (martinalderson.com)

Recent discussions in the AI and SaaS spheres reveal a significant shift in how organizations approach software procurement and development, as AI agents begin to take over roles traditionally occupied by SaaS tools. As software coding agents like Claude Code and Gemini 3 gain prominence, many software engineers are opting to build custom solutions internally rather than relying on SaaS applications, particularly for simpler tasks such as dashboards and UI/UX mockups. This trend poses a challenge to established SaaS providers, who may see declining demand and questioning of renewal quotes as companies assess whether the investment is worthwhile compared to building tailored solutions. This evolution signifies a critical turning point for the SaaS market, with potential ramifications for valuations based on customer growth and net revenue retention (NRR). As organizations begin to leverage AI agents for internal tool development, vendors may face pressures to reevaluate their pricing strategies while grappling with the complexities of maintaining their services. Although not all SaaS products are at risk—especially those requiring high uptime or regulatory compliance—the landscape is shifting, raising the bar for future SaaS offerings and exposing a growing divide between technically adept companies and those that are less capable. The rise of AI-driven solutions demands a reevaluation of what makes SaaS valuable, as companies with robust internal capabilities can increasingly compete with traditional software solutions.