Why Most SaaS Companies Won't Survive AI: The Cannibalization Problem

What's Happening

This post has been weighing on me all day. HUGE kudos to Intercom and @eoghan... legendary stuff. The numbers and turnaround are incredible, here's my tl;dr. Eoghan could do this because he had nothing left to lose. He came back from a two year hiatus to a company heading toward negative growth. That desperation removed the one thing that kills every incumbent - the instinct to protect what you already have. He killed $60M in ARR on purpose. Moved 80% of R&D to a product doing single digit percent of revenue. Replaced 'mature' board members with startup founders. Sent 100% of paid traffic to a brand that barely existed. Every late stage SaaS company has the same assets to make this shift. Brand, customers, cash flow, distribution. But assets become liabilities the second you start protecting them instead of deploying them. The reason most pre-AI software companies won't survive isn't a talent gap, it's an emotional one. Founders can't bring themselves to destroy the thing they built. And by the time the numbers force their hand, someone else already did it for them.

Why This Matters for Your Business

This is a real-world case study of the Innovator's Dilemma playing out in the AI era. Clayton Christensen wrote decades ago about how incumbents fail not because they lack resources, talent, or market position, but because their existing customers and revenue streams actively prevent them from pivoting. Intercom's CEO had the rare advantage of desperation — returning to a company heading toward negative growth gave him the freedom to torch $60M in annual recurring revenue and redirect 80% of R&D toward an unproven AI product. Most founders don't have that kind of forcing function, and that's exactly why most won't make the leap.

The practical takeaway for small and mid-size businesses is straightforward: if you're paying for SaaS tools that haven't dramatically reinvented their product in the last 12 months, those tools are likely on borrowed time. AI-native alternatives are emerging in every software category — from customer support to project management to analytics — and they're not just incrementally better, they're architecturally different. Now is the time to start evaluating replacements, not when your current vendor's product stagnates or their pricing spikes as they scramble to fund a late pivot.

For founders building new AI products, the window of opportunity against established SaaS companies is real but finite. Incumbents who do find the courage to cannibalize themselves — like Intercom — will become formidable competitors once they've shed their legacy weight. The deeper insight here transcends software: the "emotional moat" that prevents companies from disrupting themselves applies to any business transformation. The companies that thrive in the AI era will be the ones willing to grieve their old business model while it's still generating revenue. Almost nobody can do that — which is precisely why it's such a powerful competitive advantage for those who can.