Per-seat pricing was designed for companies with seats

Per-user-per-month became the default way to price business software because it works beautifully for the companies it was designed around. Value scales with headcount, procurement understands it, the vendor grows with the customer, and everybody can model it.

Applied to a firm of five, the same model does something quite different: it puts a price on hiring, charges the same rate for a person who opens the system twice a week as for one who lives in it, and makes the cheapest configuration the one where everybody shares a login — which is what actually happens.

What the model assumes

That seats correspond to value. In a large company, roughly true. In a small firm, one person uses the system all day and four touch it occasionally, but all five are billed identically.

That headcount is stable and legible. Small service firms run on part-timers, a spouse doing the books on Sundays, and a subcontractor who needs access for three weeks. None of that fits a seat.

That the customer will negotiate. Enterprise buyers do. Small firms receive an email announcing a change and decide between paying and migrating, which is not a negotiation.

That switching is possible. Also true at scale, where there is budget for a migration project. A five-person firm with three years of history in a system does not have a fortnight to move it.

Where the cost actually lands

Not on the monthly figure, which is usually survivable. It lands on decisions.

A firm that wants to give a seasonal helper access for six weeks thinks about it, because the tier jumps. A firm that would benefit from the bookkeeper seeing job data does not add them. Growing from five to six people means a licence conversation, and the ordinary consequence is that fewer people use the system than should — which quietly makes the record worse and the owner more of a bottleneck.

What the alternatives look like

Per business. One price for the firm, however many people it has. Simple, and it stops the software from having an opinion about your hiring.

Per volume of work. Charging on jobs or invoices, so cost tracks activity rather than headcount. Fairer for seasonal businesses, and it means a quiet month costs less.

Self-hosted and free. You pay for infrastructure and your own time. Genuinely cheaper for some firms and genuinely more expensive for others, and anyone who tells you it is always cheaper is selling something too.

None of these is automatically better. What matters is that the price does not attach to the thing you are trying to do more of.

Why this is a licensing question underneath

Pricing can change. A vendor that raises prices, adds a tier, or moves a feature you rely on into a higher one is doing something normal, and you have whatever leverage your size gives you.

What determines your options is the licence, not the price list. Open source does not make the software free of cost — infrastructure, time and support are all real. What it does is make the price renegotiable, because somebody else can run it, including you.

For a large company that is a nice-to-have alongside a procurement department. For a firm of five it is the only leverage available, which is why it matters more at the small end, not less.