Is SaaS dead? Not yet – But the Build-vs-Buy question CxOs are asking as already obsolete
In the mid-1990s, I was on a small Intel team building WIINGS, an indirect-inventory system for factory stores. We were told an ERP rollout would replace it within a year. Decades later, its core logic still runs, largely untouched, while the UI evolved around it.
For two decades, enterprises outsourced differentiation to SaaS in the name of speed and cost. That trade-off is breaking down, because AI has fundamentally changed the economics of building software. What once needed a large engineering teams and long software development lifecycles can now be scaffolded, tested, and refined in real time. Netflix built its moat on custom recommendation and delivery infrastructure instead of buying off-the-shelf. Shopify’s real strength is enabling deep merchant customization on top of a SaaS core. What elite engineering teams achieved at that scale is now accessible to any enterprise willing to build.
This doesn’t mean SaaS disappears, but its current form is under real pressure. The chronic CIO complaint has been paying for 100 percent of a platform while using 30 percent of its capability and inheriting a roadmap you don’t control. Salesforce has already fired the first public salvo acknowledging how much customers build on top of the core platform. SAP and Workday deployments are already heavily customized in practice. The next generation of SaaS becomes composable, AI-native, and outcome-priced rather than seat-priced; shifting from system of record to system of enablement.
Two more shifts CxOs should price into their planning now. Tech debt stops being an inevitability once AI-driven environments can continuously refactor code and patch vulnerabilities proactively. ‘Accumulated debt’ gives way to ‘continuous renewal.’ And software quality takes a real step-change. With AI generating test scenarios, simulating user behavior, and monitoring systems live, the gap between expected and actual performance narrows dramatically, meaning fewer outages and faster recovery as a direct result, not a soft benefit.
Here’s the framework that replaces ‘build vs. buy’ as the operative question – Core vs. Context. Where do you differentiate, and where do you commoditize? That’s the real decision now, not whether to build or buy. Four trends worth tracking against it. AI as an embedded collaborator across the development lifecycle, not just a tool. Hyper-personalized enterprise systems, where a two-person-plus-AI team can ship what once took a full department. Innovation harvesting at scale, moving ideas from concept to production in weeks instead of quarters, without the usual budget bureaucracy, and Digital sovereignty, enterprises reclaiming control over their own data, logic, and differentiation.
We’re coming full circle. The industry moved from custom systems to standardized SaaS in pursuit of efficiency, and AI is taking it back to custom, at a completely different scale and speed. The organizations that recognize this now will redefine their competitive advantage; the ones that don’t may find themselves constrained by the very systems that once promised to accelerate them. If you’re mid-renewal on a major SaaS contract, running it through the Core vs. Context test first is worth the delay.