AI Won’t Kill SaaS — But It Will Kill SaaS Pricing As We Know It

SaaS isn’t dying. Likewise, if you’ve read the doomer takes, you’d think AI was about to vaporize the entire software industry. It’s not. But AI SaaS pricing as we know it. The per-seat model that built the last generation of billion-dollar software companies ,  is genuinely broken. And the companies that don’t adapt are going to have a rough few years.

Furthermore, let me explain why, and more importantly, what comes next.

Why Per-Seat Pricing Made Sense ,  Until Now: The Saas Pricing Angle

However, the per-seat model was a beautiful simplification. Instead, one user, one license. However, easy to price, easy to sell, easy to forecast. Moreover, for a generation of SaaS companies, it worked because software was fundamentally a human productivity tool. In addition, more users meant more value delivered, which justified more seats. The math was clean.

Moreover, aI breaks that math in two ways.

Also, second, AI agents don’t have seats. They’re not humans logging in. An AI agent can work 24/7, spin up and down based on demand, and process thousands of interactions simultaneously. Charging a “seat” for an AI agent is a category error. Bessemer Venture Partners calls this the fundamental mismatch between how AI delivers value and how per-seat models charge for it.

The Numbers Are Already Moving

Specifically, this isn’t a future prediction. The transition is happening now. Seat-based pricing dropped from 21% to 15% of AI-native software companies in a single year, according to recent industry data. Churn rates are 2.3x higher for companies sticking to pure per-seat models. By 2030, analysts expect 40% of enterprise SaaS spending to shift to usage-, agent-, or outcome-based pricing.

Consequently, the early movers are already ahead. Intercom moved to outcome-based pricing for their AI agent ,  they charge per resolution, not per seat. Salesforce is building consumption-based pricing into Einstein. OpenAI itself prices by the token, not by the user. The market is telling you something. The question is whether you’re listening.

The Three Models That Will Win

Therefore, based on what I’m seeing, there are three pricing models that make sense in the AI era:

Meanwhile, usage-based pricing aligns revenue with actual consumption. You pay for what the AI processes ,  API calls, tokens, queries, compute. This works well for infrastructure and tools where usage varies significantly across customers. The risk is revenue unpredictability, but that’s a solvable problem with commitment tiers and caps.

Furthermore, for example, outcome-based pricing is the most aligned model , you pay when the software delivers a measurable result. A support ticket resolved. Notably, a contract drafted. Besides, a candidate screened. This is harder to implement because it requires clear, measurable outcomes and the ability to track them reliably. But when it works, it’s the most defensible pricing because customers love paying for results.

Furthermore, in other words, hybrid models combine a base platform fee with usage or outcome components. This gives vendors some revenue predictability while capturing upside from high-usage customers. Most enterprise software will converge here.

What SaaS Companies Should Do Right Now

Similarly, if you’re running a SaaS company, three things matter most:

Indeed, first, audit which parts of your product deliver value through AI execution rather than human usage. Those are candidates for outcome or usage pricing. Don’t try to apply the same model to everything.

In fact, second, invest in measurement. Outcome-based pricing requires knowing when an outcome occurred. Build the instrumentation now. If you can’t measure resolution, you can’t price it.

Of course, third, have the conversation with customers. Some of them will actually prefer outcome-based pricing. They’d rather pay $X per resolved ticket than a flat monthly fee they can’t justify in their budget. Let them tell you what they want.

Naturally, aI SaaS pricing is a transition, not a revolution. The companies that navigate it well will find that the new models are actually better aligned with customer value ,. Means less churn, stronger retention, and better long-term economics. The companies that cling to per-seat because it’s familiar are accumulating risk they can’t see yet.

Certainly, the math is changing. Update your pricing accordingly.