Scheduling software is cheap right up until the second person joins the call. After that it is not, and the reason is not the software. It is the billing model.
Per seat is the default for business tools, and for most of them it is the correct default. On a scheduling tool it does something odd: it prices the shape of your meetings rather than your use of the product. A two-person sales call, which is the most ordinary meeting shape in existence, costs twice as much per month as a one-person one. Permanently. Whether or not it ever gets booked.
What follows: that arithmetic at three team sizes, the behaviour it produces, and the honest case for the other side.
The moment the bill notices your second host
Nobody is surprised by a scheduling bill on day one. One person, one link, one small number, and it never comes up again. The surprise arrives attached to a sentence somebody says in a standup: can you be on the demo with me?
What that sentence asks for, mechanically, is a collective event type. One slot, two calendars checked, both people on the invite. On a per-seat plan the second calendar belongs to a second user, and a second user is a second licence. So the request is not really "can you join the demo". It is "can we add another seat to the subscription, indefinitely, so this meeting shape becomes legal".
The awkward part is that a licence bills continuously and a meeting does not. If the two of you run that call four times in March and not once in April, the seat costs the same in both months. It is not a price for the meeting. It is a price for the possibility of the meeting.
It is worth noticing what a second host actually costs the vendor. One more OAuth token, one more free/busy query per availability lookup, and one more row on the attendee list of an event that was going to be created anyway. The marginal load is small and roughly fixed. The marginal price is a whole seat.
The arithmetic at three team sizes
Two people
Per seat: 2 × $12 = $24 a month, or $288 a year. The second chair on its own is $144 a year, and that is the figure worth holding onto: the standing cost of one meeting shape rather than of any meeting.
Flat, on Setupp: two hosts fit inside Pro at $19 a month, $228 a year. The gap is $5 a month. Be honest about what that is: nothing. Nobody migrates a scheduling tool to save $60 a year, and if two people is your whole team then the pricing model is not your deciding factor. Choose on features and ignore this post.
Five people
Per seat: 5 × $12 = $60 a month, $720 a year. Flat: five hosts sit inside Team at $49 a month, $588 a year. The gap is $11 a month, $132 a year. Still modest, still not a reason on its own.
What has changed by five is not the size of the gap. It is the direction each line is pointing. One of them moved when you hired. The other did not notice.
Ten people
Per seat: 10 × $12 = $120 a month, $1,440 a year. Flat: ten hosts is the top of Team, still $49 a month, $588 a year. The gap is $71 a month, $852 a year, and it opened without either price being changed by anybody. Only the headcount changed.
That is the entire argument in three rows. A per-seat total is headcount multiplied by a rate, so it has no ceiling and no plateau: it is a slope, and the slope is your hiring plan. A flat total is a step function. It moves when you cross a limit and sits still in between. Between two people and ten, the per-seat line multiplies by five. The flat line steps once.
What teams do instead, and why all three are worse
The invoice is the visible cost. It is not the expensive one. The expensive one is what a team quietly starts doing to avoid the invoice, and in practice it is always one of three things.
- They avoid the meeting shape. The solutions engineer stops joining discovery calls, so the account executive relays the technical questions badly and books a second call to answer them properly.
- They share a login. Two people, one set of credentials, one seat. The invite goes out under the wrong name and the availability being checked belongs to whoever owns the account.
- They nominate an owner. One person holds the booking link and copies everybody else in. The event has one host on it, so nobody else's calendar is ever checked for conflicts.
Rank each of those against the cost it avoids and none of them survives. The second call burns more salaried time in one afternoon than the seat costs in a year. The shared login is an audit finding waiting to be written down, and it also breaks the thing you were buying, because a scheduler that cannot tell two people apart cannot check two people's calendars. The third looks fine for months: a copied-in colleague is not a calendar check, and the difference only becomes visible on the day somebody is double booked in front of a customer.
None of this is a failure of judgement by the team. It is the predictable response to a price attached to the wrong thing. Charge per head for a shared object and people will find ways to share fewer heads.
Who should actually be on a given event, and whether the work should rotate between people at all, is a real design question with a real answer, and it is covered in round-robin versus collective scheduling. The narrower point here is that a billing model should not be making that decision on your behalf.
The case for per-seat, made properly
It would be easy to write per-seat off as a trick, and it would be wrong. It is a reasonable answer to a genuine problem, and it deserves stating at full strength before anybody disagrees with it.
A vendor needs value capture that scales with value delivered. A hundred-person company gets more out of a tool than a two-person one and should pay more for it. Seats are the cleanest proxy anyone has found: easy to count, easy to audit, easy for both sides to forecast, and tied to the customer's growth rather than to metering something nobody can predict in advance.
For a great many products a seat genuinely is the unit of value. A design tool, a code editor, a CRM licence, an analytics seat: each user gets their own workspace, their own files, their own saved state and their own reason to log in on a Monday. Ten users receive ten times the product. Charging ten times for it is not extraction, it is arithmetic.
Scheduling is not shaped like that. The unit of value is the meeting, and a meeting is a shared object. When the solutions engineer joins the demo, nothing new is created for them. They are added to a booking that already exists. They do not receive their own workspace out of it, or their own availability rules, or their own pipeline. They receive a line on an attendee list and a calendar entry they did not create.
Ten users of a design tool get ten workspaces. Ten hosts on a scheduling tool get one meeting.
So the mismatch is structural rather than moral. Per-seat pricing on a scheduler charges by the number of people who could appear on an invite, inside a product whose entire job is putting several people on one invite. The model is sound. It is attached to the wrong noun.
There is a version of the counter-argument that survives all of this, and it should be said plainly: if every person on your team runs a genuinely independent booking flow, with their own event types, their own intake questions and their own pipeline, then a seat really is a workspace and per seat is measuring the right thing. Plenty of sales floors look exactly like that, and if yours does, the rest of this is not your problem.
What flat per workspace costs, and where it stops
Setupp charges for the workspace and includes hosts in the tier. The numbers in full, because a pricing argument that ends vaguely is just a pitch:
- Free, $0, 1 host. One person, one event type, every booking still scored.
- Pro, $19 a month, 3 hosts, all of them on every invite. This is the two-people-on-a-demo case.
- Team, $49 a month, 10 hosts, and the tier where collective and round-robin assignment appear.
- Business, $129 a month, 50 hosts.
Adding a host is a setting rather than a purchase order, which is why multi-host scheduling sits near the bottom of that range instead of the top of it. Nothing is metered per booking at any tier. The pricing page runs the same comparison with a slider, so you can put your own per-host rate in place of the representative one used above and trust the result more than you trust this post.
One limit on the claim, so it stays the size it should be: this does nothing for you if you are one person. One person is one seat under either model, and the comparison is empty.
The last caveat is the biggest one, and it undercuts the whole exercise slightly. The cheapest scheduling tool is the one that prevents the meetings you should not be having, at almost any monthly price. If your calendar is full and your pipeline is not, the difference between flat and per seat is a rounding error next to qualifying leads before they book. Fix which meetings happen first. Then argue about the invoice.