In partnership with
Hey friends, happy Monday!
For years, SaaS pricing had a wonderfully convenient assumption built into it: more users usually meant more value. A company hired more people, those people needed software, and the vendor sold more seats. Simple enough.
AI is starting to break that relationship, because the better the software gets at doing the work itself, the fewer human users a customer may actually need.
Imagine a 50-person support team using an AI agent to handle the same workload with 20 people. The customer is getting more value from the product, but under traditional per-seat pricing, the vendor could end up making less money precisely because its product is working so well.
And that is where things get interesting.
As AI moves from assisting people to completing more of the work itself, pricing has to move beyond who uses the software and get closer to what the software actually accomplishes. That might mean charging for usage, credits, completed work, outcomes, or some combination of them.
There is no obvious winner yet, and there probably shouldn't be, because an AI copilot and an autonomous support agent create value in very different ways.
Which is why this week's edition isn't really about finding the perfect AI pricing model. It's about figuring out what your customer is actually paying you to accomplish, and then building your packaging and pricing around that.
In today's edition, we’ll explore:
Why traditional per-seat pricing starts to break as AI does more of the work
How seats, usage, credits, and outcomes change who carries the economic risk
Why tokens make sense as a cost metric but rarely as a customer value metric
How leading AI companies are experimenting with new packaging and pricing models
A practical framework for finding the right value metric for your product
A pricing audit and 90-day workflow for testing a new model
The big idea is simple: don't price the AI itself. Price the unit of value the AI creates.
— Naseema Perveen
Subscriber exclusives: Discover what’s behind AI headlines
THE CONTROL LAYER WITH VICTOR DEY
The real decisions shaping AI are rarely made in public. In The Control Layer, Victor Dey speaks directly with the engineers, executives, operators, and skeptics shaping the industry. Discover the tensions behind AI’s biggest stories, from ambition and guardrails to trust, control, and who ultimately gets to decide how AI is used.
JOIN THE CONVERSATION SHAPING THE GLOBAL AI NETWORK
The AI industry is growing quickly, but the standards, professional support, and shared best practices around it are still taking shape. The Global AI Network (GAIN) is being created to help close that gap by building a trusted professional community for people working across AI.
We are gathering feedback ahead of the launch, and your perspective can help shape what GAIN offers its members. The survey takes just two minutes, and your response can remain anonymous.
Complete the survey and help shape GAIN.
