SaaS is about to lose its moat

The software-as-a-service model has been the dominant force in enterprise technology for twenty years. It delivered predictable revenue, strong margins, and the kind of recurring subscription income that investors have rewarded generously.

That model is not collapsing. But the structural advantage it has always relied on, the idea that switching costs make customers sticky, is quietly being eroded. Most valuations have not caught up with that reality.

What made SaaS defensible

SaaS products built their moats in a specific way. Implementation was expensive and time-consuming. Migrating data was painful. Retraining staff was disruptive. And building an alternative yourself was not a realistic option for most businesses.

The combination of those factors kept renewal rates high and churn low, even when customers were unhappy. Staying was less painful than leaving.

This was genuinely a strong structural position. The best SaaS businesses compounded on it: deeper integrations, wider API ecosystems, data network effects. The more embedded the product became, the harder it was to remove.

Where the erosion starts

AI is changing the build side of that equation.

The cost of creating a bespoke system that replicates 80% of what a mid-tier SaaS product does is falling rapidly. What previously required a team of developers, a long project timeline, and significant ongoing maintenance can now be delivered by a single technically capable person in a fraction of the time.

That does not make every SaaS product immediately replaceable. But it does mean that the barrier to walking away is lower than it has ever been.

When the cost of a bespoke replacement becomes comparable to a few months of SaaS subscription fees, the economics of the lock-in model start to shift.

Not all SaaS is equally exposed

Deep compliance and integration complexity provides genuine protection. A payroll platform embedded in HMRC reporting workflows, or a medical device management system that requires regulatory certification, has real switching costs that AI tooling cannot eliminate overnight.

Generic horizontal platforms are more exposed. CRM systems, project management tools, basic HR platforms: these serve broad markets with broadly similar needs. A competent consultant with AI assistance can replicate the core functionality for most small or mid-size businesses in a reasonable timeframe.

The market will not feel this immediately. Enterprise procurement cycles are slow. Budget holders are cautious. And there is a credibility gap: most businesses do not yet believe that custom software can be built and maintained at a cost that would make it worth seriously considering.

That credibility gap will close. It is already closing for the businesses paying attention.

What the fragmentation looks like

The most likely outcome is not that SaaS disappears. It is that the market fragments.

Large enterprise platforms with deep integrations and compliance requirements will retain their position. They have earned their moat and the switching costs remain real.

The middle tier, the platforms serving small and mid-size businesses with generic functionality, faces real pressure. Not immediately, and not from a single disruptive competitor. From a long tail of smaller operators, each serving twenty or fifty customers with a product tailored to a specific context, at a cost point that starts to look competitive.

This is not a new business model. It is closer to how bespoke software worked before SaaS existed. The difference is that AI has dramatically lowered the cost of building and maintaining it.

The question worth asking about valuations

For anyone evaluating SaaS businesses as investments, the question worth asking is not whether the product is good. It is whether the switching costs are real, or whether they are a legacy assumption from a world where building alternatives was genuinely difficult.

Many SaaS valuations are built on the second type. Sticky customers, low churn, predictable revenue: all of those numbers look good right up until the moment that building a replacement becomes a straightforward consultancy engagement rather than a multi-year project.

The numbers will not move suddenly. But the underlying assumption is shifting. That is worth pricing in before the market does it for you.