SaaSmageddon Is Here – and Not All Software Stocks Will Survive

Welcome to the end of the world for Software-as-a-Service (SaaS) stocks – at least as we’ve known them. 

Recently, we’ve been watching those trades bleed out like a secondary character in a Tarantino film. As a group, they have dropped more than 20% since late 2025 – one of the fastest drawdowns for the SaaS cohort outside of the 2022 tech unwind and the 2008 financial crisis.

But here is the twist: This isn’t a macro problem. Unlike prior tech selloffs, this one isn’t being driven by tightening financial conditions or collapsing demand.

What’s happening now is bigger than all of that. 

This is a displacement event – and we’re calling it SaaSmageddon. 

Why AI Is Collapsing Seat-Based SaaS Models

For 15 years, the Software-as-a-Service business model was like the ultimate profit hack. Build an attractive dashboard, connect it to a database, and charge companies $30 to $100 per month, per human, to use it. 

The more workers those client companies hired, the more money SaaS providers made.

But now, the market has realized that AI is beginning to erode the human input in that cash equation – and it’s happening faster than most investors expected.

Unlike copilots that assist a user, agentic systems like Anthropic‘s Claude Cowork are designed to complete multi-step workflows autonomously – from research to execution – without persistent human supervision. And when an AI agent can perform the work of five junior analysts or paralegals, the enterprise doesn’t just need fewer employees – it needs fewer software licenses. The “seat-based” moat is being drained.

There are really three critical dynamics at play here.

The Flattening of the Middle Layer

Most horizontal SaaS companies function as expensive intermediaries between a human user and a structured database. Agent frameworks using tools like Model Context Protocol (MCP), function calling, and API-native execution increasingly allow models to retrieve, update, and reason over data without a graphical interface.

Project Genie and the Death of Creation Moats

Google’s Project Genie has turned high-fidelity creation – whether that’s a video game or complex UI – into a prompt. The “barrier to entry” for many categories of application software is collapsing. When small teams can generate high-quality apps in days instead of months, why pay a premium for legacy tools?

The “Show Me the Money” Pivot

After three years of AI hype, investors have stopped asking, “What is your AI strategy?” and started asking, “Where are the profits?” While chipmakers like Nvidia (NVDA) and Micron (MU) are printing money, software giants like Salesforce (CRM) and Adobe (ADBE) are seeing their multiples compress as they struggle to monetize AI without cannibalizing their own seat-count revenue. In effect, AI is expanding margins at the infrastructure layer while compressing them at the application layer.

In other words, we think it’s inevitable that AI will flatten and obsolete the middle software layer that so many SaaS stocks thrived in for 10-plus years. 

That doesn’t mean all software stocks are doomed. But, frankly, most are…