PagerDuty ditches per-seat pricing as AI rewires its business

Last week at HumanX, I heard a lot about AI, some hype, and some more practical, but however it's presented, it's undeniable that AI is driving changes across every company and industry. I spoke to PagerDuty CEO Jennifer Tejada, who told me how it’s forcing her company to rethink systems failure, how customers react to incidents and even how to price the product.

PagerDuty gets its name from the old practice of putting IT staff on call with physical pagers so they could fix critical systems in the middle of the night. Its customer base includes over two-thirds of the Fortune 100 and half the Fortune 500, and when systems go down at that scale, it is not just an annoyance, it tends to cost a lot of money, anywhere from $300,000 to $1 million or more an hour, depending on the type of business.

That means it is not only crucial to discover the root of the issue and get back to normal operations as quickly as possible, it is also critical to understand what happened to try and prevent that same type of failure from happening again.

This is particularly important because 70% of major incidents on the platform repeat themselves, according to Tejada, not because teams aren't capable but because the systems are complex. "Resilience is about as you solve for that failure, what did you learn from it? And how did you put yourself in a less fragile or stronger, more resilient position afterwards," she said.

How AI changes everything

Customers may be only starting their AI journey, but vendors like PagerDuty need to be ahead of them. "[Many of our customers] are only part way along in their cloud adoption cycle, and now they're learning and finding ways to leverage AI, but also having to manage a lot more risk than they have in the past."

Tejada says, it not only is a new way of working, it also presents a new way of failing, which in many ways makes it harder to detect and treat. Whereas when traditional software stops working, you know it's broken, with AI it's more subtle. 

"In fact, it's harder because when something goes down you actually know it's broken, but when AI drifts you can't see it right away. When an agent breaks, you don't know until it has broken other things in its path," Tejada said. She said PagerDuty is focused on catching small problems before they snowball into major incidents. "The goal is to shift left to prevent those small incidents from becoming major material business impacts," she said.

New ways of measuring value

If AI makes it more challenging to detect incidents, measuring value by the number of seats makes even less sense. That change is forcing the company to rethink the way it charges customers for its services, and as such it's moving away from the per-seat license model of the past to a more outcome-based approach. Tejada believes that every company wants a clear understanding of where the value is being created.

"In our business in particular, because we have built so much automation in the platform, having a seat-based license model is not the best metric to demonstrate value," she said.

"We have a customer who says every hour costs them a million dollars during a major incident. And if we can help them compress a major incident from lasting two hours to 20 minutes, that's a massive value. Well, that's a performance-driven use case," she said. PagerDuty just has to figure out how to put a dollar value on that savings.

Part of this is a creative reaction to a changing market. As companies cut back on their IT staffing, it has coincided with buying fewer licenses. The shift to usage‑based, value‑aligned pricing is both a response to that compression and to new AI use cases, where pricing on consumption makes more sense given the high cost and complexity of delivering AI solutions.

One of the issues for CFOs in moving to usage-based is lack of cost certainty, but Tejada says that CFOs she's been talking to see the advantage. "They're pretty savvy at managing usage-based models, and they want to be sure that the usage is connected to value creation somewhere in the business, whether that's cost avoidance, hard cost savings, revenue capture, demand growth, competitive advantage, etc," she said. 

That's really what any business is looking for when it comes to buying software, but McKinsey's most recent State of AI report found that while nearly 90% of organizations are using AI, only 39% can point to any bottom-line impact, making finding and fixing problems even more imperative.

Tejada says every usage metric on the platform has been growing, and the new pricing model is designed to capture that. "There's definitely more demand on the platform, and with the shift towards [usage-based] pricing, we start to capture more of that demand through usage, as opposed to just how many people do you need licensed to the platform," she said.