Opinion
“SaaSmargeddon” is here: AI threatens the core of Software-as-a-Service
AI is rewriting the rules of SaaS, forcing companies to rethink pricing, moats, and data lock-ins.
The term "Saasmargeddon" (The SaaS Armageddon) perfectly captures the existential dread currently rippling through Silicon Valley and the broader tech industry.
As software companies' stocks are facing significant declines, some of their CEOs are being "pushed" towards resigning for "not preparing for the AI disaster in due time."
The SaaSmargeddon has just started to take its toll.
To understand the reasoning behind the current situation, there are a few areas where AI challenges the SaaS model.
1. The Death of the "Seat-Based" Model
For twenty years, SaaS grew by charging "per head." But if Gen AI allows one person to do the work of five, or if an AI agent replaces a human role entirely, the customer no longer needs those extra seats.
The Fear: Revenue contracts even as customer productivity explodes.
The Pivot: Companies like Salesforce and Intercom are scrambling to move toward outcome-based pricing (e.g., "pay per resolved customer issue" rather than "pay per support agent").
2. The Collapse of the "Moat"
Traditionally, a SaaS company's "moat" was its complex code and user interface. Today, "Vibe Coding" (using natural language to generate apps) enables a small team to replicate a "point solution" (such as a simple e-signature or HR tool) in a weekend.
The Fear: If software is "cheap" to build, it becomes a commodity with zero pricing power.
The Pivot: SaaS must move from being a "system of record" (where data is stored) to a "system of intelligence" (where AI actually utilizes data to drive decisions).
3. Data Gravity vs. Data Portability
In the past, getting your data out of a SaaS platform was a nightmare, which kept customers "locked in." AI agents can now "read" and "migrate" data across platforms with ease.
The Fear: High switching costs, the bedrock of SaaS valuations, are vanishing.
4. The Rise of "Agentic" Workflows
We are moving from "Software as a Service" to "Service as Software." Instead of a human logging into a dashboard to click buttons, an AI agent will run in the background. If there is no human user, do you even need a UI?
The Fear: The "headless" software era could make many famous front-end interfaces obsolete.
The above stated the challenges, but not the solutions. The reason is that solutions that maintain 8-10 billion in ARR (Annual Recurring Revenue) are yet to be found.
Avihai Michaeli is a senior investment banker, and served as a Partner at PwC.
