Software stocks shed around $2 trillion in market capitalisation in early 2026. Analysts called it the “SaaSPocalypse.” If you’re building a SaaS business and thinking about raising, you’ve probably seen the coverage and wondered whether it’s something to worry about.

The short answer: it depends on what you’re building. Here’s the longer one.

In this article, we cover what caused the selloff, why it’s mostly a large-cap story, how smart SaaS companies are responding, and what it means for investor expectations in 2026.

What actually happened

The selloff wasn’t triggered by bad earnings or a macro shock. It was a structural reassessment.

Investors looked at the rise of AI agents and asked a simple question: if AI can now do the work of the humans using your software, what happens to your per-seat revenue? Per-seat pricing has been the engine of SaaS growth for two decades – you sell licences, customers grow their teams, they buy more licences. AI disrupts that assumption at the base level.

For large incumbents, the answer was uncomfortable. Enterprises are reducing software licences rather than growing them, as AI-enhanced workers accomplish more with fewer tools. Salesforce fell 28%. Adobe dropped more than 25%. Workday announced layoffs attributed to AI efficiency gains at its own customers – meaning customers were doing more with fewer people, and buying fewer seats as a result.

These are businesses with billions in per-seat revenue and customers who now have genuine alternatives. That’s a real problem – but it’s not necessarily your problem.

Why this mostly doesn’t apply to early-stage founders

If you’re building anywhere from pre-seed through to Series A, the SaaSPocalypse is largely a large-cap story.

The incumbents taking the hardest hits are selling commoditised, workflow-layer software where AI agents can credibly replicate the core functionality. Gartner estimates that by 2030, around 35% of point-solution SaaS tools will be replaced or absorbed by AI. The other 65% will survive and grow.

The businesses that struggle will be those with generic, undifferentiated tooling and no real data moat. The businesses that thrive will be those solving specific, complex problems in industries where domain knowledge matters – where the value isn’t just in the software, but in what the software knows.

If your product does something genuinely hard in a specific context, whether that’s healthcare, construction, agriculture or anything else with real operational complexity, the SaaSPocalypse is mostly background noise at this stage.

What it does affect is investor psychology. The era of funding growth at almost any cost is over. Investors want to see unit economics, efficiency and a clear path to profitability – which means understanding your metrics is more important in 2026 than it was in 2021, not less.

The more interesting story: AI inside SaaS products

While the headlines focus on AI eating SaaS from the outside, a lot of SaaS companies are embedding AI directly into their own platforms – and the results are genuinely interesting.

HubSpot has embedded AI across its CRM to surface predictive lead scoring, deal insights and automated content generation – not as bolt-on features, but as core to the product experience. Xero, through its AI layer Jax, is turning transactional accounting data into forward-looking cash flow intelligence directly within the workflow. The AI makes these products stickier, not more vulnerable.

The most instructive example is Procore, the construction management platform. It launched Procore Helix – an intelligence layer built on data from over three million construction projects. Helix analyses daily logs and submittals to surface project risk before it becomes a problem, predicts schedule impacts and benchmarks performance against real industry data. No generic AI model can replicate that from the outside, because it’s built on proprietary data that took years to accumulate.

That’s the key insight for founders thinking about where AI fits in their own product. The technology is increasingly commoditised – you can access powerful models via API for very little. What isn’t commoditised is domain-specific knowledge. The data your product accumulates, the patterns it learns, the proprietary signals it can surface – that’s the moat.

What this means for valuations and pricing

Valuation recovery in SaaS is coming, but selectively. The companies that will command premium multiples again are those where AI inside the product creates measurable, defensible value that customers can’t easily replicate themselves. Not across the board – for businesses where the intelligence layer is genuinely part of what makes the product worth paying for.

Usage-based pricing is part of this shift too. As AI moves SaaS from a tool you access to a service that actively delivers outcomes, per-seat pricing starts to make less sense. Pricing tied to usage, results or consumption aligns better with how value is created – and it’s what investors increasingly want to see. If your product is delivering outcomes and you’re still charging by seat, it’s worth asking whether your pricing model is telling the right story.

What to take away

The SaaSPocalypse is less a threat and more a design brief. Build with domain-specific intelligence at the core. Think about what proprietary insights your product can surface that customers can’t get anywhere else. Consider whether your pricing model captures value from outcomes rather than headcount.

And make sure you know your numbers. Investors in 2026 aren’t less interested in SaaS – they’re more discerning about which SaaS businesses they back. The ones that get funded will be those that can tell a clear, evidence-based story about efficient, defensible growth.

That story starts with your metrics.

Make sure your metrics are investor-ready

Disclaimer: This article is for general informational purposes only and does not constitute financial, legal or tax advice. Please speak with a qualified adviser (hey, that’s us!) before making decisions based on your specific circumstances.