In February 2026, financial markets did something unusual: they erased nearly $2 trillion in value from software companies in just a few weeks. There was no fraud. No banking crisis. There was a question that investors could no longer ignore:
If an artificial intelligence agent can do the same work as ten employees using that software, why keep paying ten monthly licenses?
That breaking point became known as the SaaSpocalypse.
What is the SaaS model and why does this matter?
SaaS stands for Software as a Service. It's the model behind tools used every day in businesses: corporate email, customer CRM, support ticketing systems, HR management platforms.
The logic was simple: instead of buying software once, companies pay a monthly or annual subscription. And that subscription is charged per person, per "seat." 10 employees using the tool: 10 seats. 500 employees: 500 seats. Revenue was predictable, it grew alongside the company, and canceling was complicated enough that almost no one bothered.
That model worked flawlessly for over twenty years. Until AI agents arrived and emptied the seats.
The seat that went empty
An AI agent is a program that can execute complex tasks on its own, without anyone sitting in front of a screen. It can reply to emails, classify documents, update records, process requests and generate reports: everything a human employee used to do with their session open in the software.
When that happens, licenses become redundant. A company that once needed 50 seats for its CRM may only need 15, with agents doing the work of the other 35.
During the 2025 renewal cycles, large corporations began not renewing. Technology teams started auditing their software stacks with a new question: can the AI agent we already have do this?
Investors noticed. And they reacted.
The companies that felt it hardest
This table shows how valuations changed for the top SaaS companies before and after the SaaSpocalypse:
| Company | What it does | Historical multiple | 2026 multiple | Stock decline |
|---|---|---|---|---|
| Salesforce | CRM / sales | 31x | 11x | -30% |
| Adobe | Design / creativity | 27x | 8x | -42% |
| Workday | HR management | 34x | 12x | -33% |
| Intuit | Accounting / taxes | 34x | 11x | -11% in one day |
| ServiceNow | IT automation | 39x | 24x | -15% in one day |
(The multiple indicates how much investors pay for each dollar of expected earnings. A lower number means less confidence in future growth.)
The most striking figure is Adobe: it went from trading at 27 times its earnings to just 8 times. Not because its results were bad (revenue grew 13% in 2026). The market simply stopped believing that growth would hold.
The problem isn't that the software stopped working
Here's the part that confuses people: most of these companies are still growing. ServiceNow grew revenue 22% in Q1 2026. Adobe posted record revenue. Customers didn't cancel all at once.
The market didn't punish today's results. It punished uncertainty about tomorrow's results.
The market didn't panic over current numbers. It panicked because the model that generated those numbers no longer seems guaranteed.
What comes next: three trends that will define the future
1. Per-seat pricing is going extinct
The model of charging per person has a limited future for most tools. New models charge by outcome: per contract reviewed, per ticket resolved, per report generated. Goldman Sachs calls it "Results-as-a-Service." ServiceNow already generates 50% of its new contracts through consumption-based models, not per seat. Companies that don't make this shift will lose customers at every renewal cycle.
2. Platforms win, niche tools disappear
The average number of SaaS apps per company dropped from 289 in 2024 to 254 in 2025, and keeps falling. Organizations are consolidating: they prefer one large platform that solves many problems over ten point tools that each solve just one. Companies consolidating their stacks are reporting cost reductions of 20% to 35%. The tools that survive are those deeply embedded in critical workflows.
3. Proprietary data is the new power
The most valuable asset in the age of agents isn't the software's feature set: it's the data that software holds. A platform that integrates identity, communication, documents, automation and business data in one place has an advantage no external agent can easily replicate. That's exactly what Microsoft 365 does: an ecosystem where the agent works on the organization's own data, without needing to connect ten separate tools.
The tools aren't dying. The business models are.
The SaaSpocalypse won't kill enterprise software. It will select it. Tools that existed only because canceling them was inconvenient will disappear. Those with proprietary data, deep integration and the ability to demonstrate concrete value will come out stronger.
For organizations, this is a real opportunity: for the first time in years, they have negotiating power with their software vendors. And they have a valid reason to ask which tools actually deserve a place in their stack next year.
If you want to keep reading about these trends before they hit the headlines, subscribe to the macareno.net newsletter. No noise, just what matters.
Sources
- $300 Billion Evaporated. The SaaSpocalypse Has Begun. — Forbes
- Are NOW, CRM, WDAY, and ADBE Stocks Bargains After Their Valuation Reset? — Invezz
- AI Agents Just Erased $2T in SaaS Value — Who Survives — Tech Insider
- ServiceNow Stock Plunges 15% as SaaSpocalypse Debate Reignites — Salesforce Ben
- Platform Consolidation in 2026: Why Companies Are Reducing Their SaaS Stack — Vantagepoint