It quietly decides who is allowed to see the truth. Every seat is a small decision about whether one more person gets access, and the predictable answers are a shared login, an exported spreadsheet, and a record that is no longer complete.
This is not an argument about price. Per-seat pricing can work out perfectly reasonable on the invoice and still change the behaviour of the team in ways nobody chose.
The mechanism
A system of record is worth exactly as much as the record is complete.
Per-seat pricing makes completeness cost money, one person at a time. And the people at the margin are never the sales team, whose seats are obviously justified. They are the ones at the edges: the delivery lead who knows this client is unhappy, the finance person who knows the last invoice went unpaid for ninety days, the support manager who has had four tickets from them this month.
Each of those people holds context that changes how an account should be handled. None of them will clear the bar for a seat.
What teams do instead
| The workaround | Why it is rational | What it costs |
|---|---|---|
| Share a login | Free, immediate | No audit trail, and nobody knows who changed what |
| Export to a spreadsheet | Everyone can read it | The spreadsheet gets edited, and now there are two records |
| Route through a gatekeeper | Keeps the licence count honest | Updates happen when that person has time, which is weekly at best |
| Just do not record it | Costs nothing today | The thing most likely to happen, and invisible until it matters |
The last row is the one that does the damage, because there is no artefact to find. Nobody files a ticket saying the record is incomplete. You discover it when a rep calls a client the week after a delivery escalation nobody told them about.
The second-order effect on reporting
Once the record is partial, every number built on it inherits the gap.
Pipeline looks healthier than it is, because the accounts with problems are the ones whose problems live outside the system. Win rates by source are wrong, because the deals that died for delivery reasons are recorded as lost on price. The forecast is assembled from the view of the people who had seats, which is the view with the least bad news in it.
None of this reads as a pricing problem when it surfaces. It reads as a data quality problem, and teams go looking for a data quality solution.
When per-seat is the honest model
It should be said plainly: per-seat is not a trick. For plenty of products it reflects genuine cost, and it keeps the entry price low for a team of three, which is a real benefit for a business starting out.
The mismatch is specific. It is bad for systems of record, where value depends on completeness and the pricing taxes exactly that. It is fine for products where each user genuinely consumes something.
The test
Not what you pay. This:
Does adding one more person to the system require a purchasing conversation?
If yes, your record will be shaped by who could afford a seat rather than by who knows something. If no, you at least have a chance at a complete picture.
Empiraa Signal prices by plan with unlimited users on every one, which is the direct answer to this problem and the reason it is built that way. The full argument for it is on unlimited users, and what Signal costs has the plans.
Related reading: what a free CRM actually costs you, and is your CRM costing you more than you think.
Compare Signal with HubSpot and Salesforce, with list prices and the date they were checked.
