What Is Per-User (Seat) Licensing?

Per-user (seat) licensing prices software by the number of named users: each person who needs access counts as one paid seat, regardless of how many devices they sign in from. It is the dominant model for subscription software, and it makes the bill a direct function of headcount - which is why a tidy seat count is one of the easiest cost savings in IT asset management.

How seat licensing works

A seat is bound to a named identity, usually an email address or directory account. That person can use the software on their laptop, phone, and home machine under the same seat, because the license follows the user rather than the hardware. Pricing is typically per seat per month, often with tiers, and the vendor’s terms almost always prohibit shared logins - one seat, one human.

The model’s strength is that it scales naturally with the team: hire someone, add a seat. Its weakness is the mirror image: when someone leaves or stops using the tool, nothing forces anyone to remove the seat, so the count only ever drifts upwards.

Per-user vs per-device licensing

Per-device licensing attaches the right to a machine instead of a person. The choice follows the usage pattern:

• One person, several devices - per-user wins; you pay once for the human, not three times for their hardware.
• Several people, one device - per-device wins; a shared warehouse workstation, lab PC, or front-desk machine serves a rotating cast for a single license.
• Concurrent licensing is the third option: a pool of floating licenses sized to peak simultaneous use, common for expensive specialist software used occasionally by many.

Most organisations end up with a mix, which is fine - as long as the register records which model each product uses, because each one drifts in a different way.

Where seat counts drift

• Leavers keep their seats. Offboarding collects the laptop but nobody cancels the logins, so the seat keeps billing for an empty chair.
• Duplicate accounts - the same person under two email addresses after a rename or a department move.
• Shadow purchases - a team buys its own seats on a card, outside any review.
• “Just in case” provisioning - seats added for a project that ended, contractors who finished, or new starters who never activated.

Each seat is small; the drift is not. Because adding a seat is self-service and removing one requires someone to notice, the gap between seats paid and seats used widens until somebody reconciles the two.

Keeping seat counts accurate

The remedies are procedural, not technical: give every per-user product a named owner, reconcile paid seats against active users before each renewal rather than after, and make seat removal an explicit step in the leaver process - the same checklist moment as collecting hardware at the end of the IT asset lifecycle. In AMPthilly, licenses and seats sit in the same register as the equipment they are used with, and offboarding checklists give the “reclaim seats” step a named owner and a due date instead of leaving it to memory.

Related terms

• ITAM (IT Asset Management) - the discipline seat tracking belongs to
• IT Asset Lifecycle - the joiner-to-leaver path every seat should follow
• Hardware Refresh Cycle - device replacement, the hardware twin of seat reviews
• Golden Image - the standard build that decides which licensed software ships on every machine
• Asset Agent - automated inventory of what is installed versus what is paid for