AI Agents and SaaS Pricing: What Happens When Nobody Churns and Revenue Drops Anyway
If you sell software per seat, this one is for you, and it is probably a few months overdue.
Nobody churns. Your customers have never liked your product more. Your revenue drops anyway.
That is what AI agents and SaaS pricing can look like when they meet:
a happy customer,
a healthy relationship,
a smaller invoice.
It does not arrive as a cancellation email. It arrives as a renewal with fewer seats, usually on a call where procurement is already present.
How AI Agents Change SaaS Pricing Before Anyone Churns
Per-seat pricing grows when your customer hires. The company adds people, more people log in, and your revenue rises with their headcount. Right now, many of your customers are rolling out agents instead of hiring. An agent does what three people used to do, and three seats quietly disappear at the next renewal.
If those agents are yours, it gets stranger. The better your AI features become, the fewer humans your customer needs, so you end up being punished for the product working, which is a strange way to run a company.
The second squeeze nobody talks about
There is another half to this problem. Every task your product performs burns compute, so your costs rise with the work being done while your price stays flat per human. I cannot see that gap closing by itself.
Which SaaS Products Are Most Exposed to AI Agents
A few months ago I read a quote that made me put my phone down. Eran Zinman, co-CEO of monday.com, said more or less that nobody is going to buy software that is not doing most of the work for them. monday.com is a public company with over a billion in revenue, and it sells seats, just like most of the founders I work with. So the person who built one of the biggest seat businesses on the planet told you, on the record, that what you sell is the old thing. I found that impressive and a bit unsettling in about equal measure.
To be fair, not every product gets hit the same way. Atlassian’s seats are still growing, apparently because agents in Jira give people more work to track. That makes the most useful question a simple one: what does your product do all day? If it is where people coordinate, seats might hold for a while. If it does the work, you are selling seats to a customer who is trying to need fewer of them.
Why Founders Wait on SaaS Pricing Changes
Most CEOs I talk to can see this coming and are waiting for renewals to force the issue. I get it. Repricing is a shit job. ARR goes wobbly, the forecast gets harder to defend, sales comp is built on seats, and someone has to call your best customers to explain that the way they pay you is changing. That someone is usually you.
But think about what renewal looks like from their side. They have just rolled out agents, there is a list of users who have not logged in since March, procurement is on the call, and they get to decide what you are worth on their timeline. You agreed to that haircut by waiting. Avoidance feels kind, and it costs you more with every quarter you leave it.
What to Do About AI Agents and SaaS Pricing This Quarter
Start with the ugly numbers
Take your top ten accounts, assume each one cuts a third of its seats, and look at your ARR eighteen months out. Write the number down. It is much less scary on paper now than it will be on a Sunday afternoon when your thoughts drift into anxiety.
Pick a direction
Choose between usage, outcomes, credits, or a hybrid with a platform fee underneath. It does not need to be the finished model. “Platform fee plus credits over the next three renewals” is a strategy. “We’re keeping an eye on it” is a hope with a slide.
Tell your team before your board
Your team needs to hear the plan before the board does. Sales especially, because their pay is built on seats and they will fill any silence with the worst version they can imagine. Give them the plan and a date for the next update, then keep that date.
Put it on the board agenda yourself
Add it to the next meeting, covering your exposure, where you are heading, and whatever you need from the board to get there. The conversation is coming either way, and the same facts land very differently depending on who says them first.
Zinman said it after his stock had already taken the hit. You have the chance to say it first, while your numbers still look decent.
If you want someone to pressure-test the numbers or the board story before your investors see it, that is work I do with CEOs. Otherwise, tonight, top ten accounts. Write the number down and see how it sits.
What’s Next?
Join My Weekly Newsletter! Get founder insights, funding strategies, and exclusive resources straight to your inbox.
Want more insights on building a thriving startup without burnout?
