The Future of SaaS in the AI Era
In the age of AI, traditional Software/SaaS are at risk and some experts are suggesting that many traditional SaaS companies will die. However, I don’t think that SaaS is going to get fully replaced by AI agents and AI workflows. There will be a transformation and an evolution of what SaaS will look like, not a demise/extinction – but SaaS is going to need to rapidly evolve. SaaS companies with strong moats, sticky data, and high switching costs will remain highly defensible against pure AI disruption.
The underlying architecture of SaaS is shifting, with intelligence moving from application layers to AI agents that orchestrate across systems. Legacy SaaS won’t disappear overnight, but the value proposition is moving from the application interface to the AI orchestration layer.
But the SaaS landscape is shifting in meaningful ways:
- Budget reallocation, not replacement. AI is shifting corporate IT spend rather than fully replacing products. Enterprises are reallocating toward AI agents, workflows, and new capabilities. This pressures existing SaaS contracts without eliminating them.
- Legacy interfaces feel dated. Static dashboards, manual processes, and non-AI-native UX look increasingly outdated compared to modern alternatives. Incumbents must adapt quickly or lose relevance.
- CIO priorities favor consolidation. Companies want fewer vendors, integrated platforms over best-of-breed point solutions, and reduced complexity. This benefits entrenched players who can add AI features more than it helps new AI startups fighting for a seat at the table.
- Model improvement changes everything. What seemed impossible a year ago is now reality. AI-powered SDR teams are replacing human ones. SaaS vendors must continuously reinvent or risk disruption as capabilities evolve and enterprise security requirements catch up.
I remain optimistic for durable SaaS businesses that will integrate AI effectively – but standing still today is not an option and SaaS companies need to rapidly evolve and implement an AI strategy.
AI will shift from tools that assist humans to systems that execute work, which compresses many feature-level SaaS products and enables AI agents to replace narrow point solutions. The SaaS market will stratify into 3 groups: weak point solutions disappear, moderately differentiated tools survive by embedding AI and changing pricing models, and AI-native companies plus deeply embedded platforms with data and workflow control thrive. Durable SaaS advantages will come from proprietary data, ownership of enterprise workflows, deep integrations, and systems-of-record positions, while simple UI-based or feature-level products without these moats become structurally vulnerable.
Also – see my original post from this summer (after AI Code was made available broadly in May 2025):