The default way software has been priced for years is per seat — pay for each user who logs in. It's familiar and simple. It's also increasingly a poor fit for how software actually creates value, and the model is starting to break down. The reasons why say a lot about where software is heading.
Why per-seat made sense — and why it punishes growth
Per-seat pricing assumes software is a tool that humans operate, so charging per human is a rough proxy for value. Fair enough, once. But it has an ugly side effect explored earlier in this series: it taxes you for growing. Add a warehouse hire who needs to check one thing, pay another full seat. The pricing punishes you precisely for adding the people that signal success, which puts the vendor's interest at odds with your growth.
What AI breaks
AI snaps the already-strained logic. When the software itself does work — drafts the follow-ups, answers the questions, generates the forecasts — that value doesn't map to a human seat at all. Who's the "user" when the system did the task? Per-seat pricing can't capture value that isn't tied to a person logging in, because the work increasingly isn't done by people logging in. The unit the model was built on is dissolving.
Where pricing is heading
As seats stop mapping to value, pricing is shifting toward what the software actually delivers — the business it runs, the outcomes it produces, the scale of operation it supports — rather than how many people touch it. The details vary, but the direction is consistent: away from a per-human tax that penalizes growth and ignores AI-done work, toward models that let your whole team use the system and that don't make you flinch every time you add a person.
For buyers, this is good news worth watching for. Pricing that charges per seat increasingly reflects an outdated idea of what software is. The tools aligning with where things are headed let everyone use the system without per-head penalties — which means the software's interest is aligned with your growth instead of taxing it.