A flat line against a sloped one
Per-seat pricing is a horizontal line: seats multiplied by a rate, independent of how hard anyone works the product. Consumption pricing is a two-part tariff — a fixed platform fee plus a rate per unit — which draws a sloped line starting above zero. Two straight lines, one of them flat, cross at most once. Everything else in this comparison follows from where you sit relative to that point.
The two models also differ in something the arithmetic does not show: who carries the variance. On seats, the vendor carries it. Your bill is the same in a quiet month and a launch month, and the vendor absorbs the cost of the extra load. On consumption, you carry it. A viral week, a runaway integration or a badly written retry loop lands directly on your invoice.
That is why the peak-month figure in this calculator matters as much as the expected-month one. A consumption plan that saves 15% on a median month and costs 120% more in a peak month is not obviously the better deal, particularly if the budget is annual and fixed. Read both numbers before choosing.
The third number worth carrying into a negotiation is the effective unit price on the seat plan: what you are paying per unit of actual work when the bill is fixed. It converts a seat quote into the same currency as a consumption quote and often reveals that a heavily used seat plan is already extremely cheap per unit — or that a lightly used one is not.
Where the crossover comes from, and when it does not exist
Set the two plans equal. The seat plan costs n × s. The usage plan costs F + (U − U_inc) × r for any volume above the allowance. Solving for U gives
U* = U_inc + (n·s − F) / r.
Read the numerator first. It is the seat plan minus the platform fee — the fixed-cost advantage the consumption plan starts with. Divide by the unit rate and you get how many units that advantage buys, then add the allowance because those units were free anyway.
Two cases produce no crossover at all, and both are real answers rather than errors. If the platform fee alone exceeds the seat plan, the numerator is negative and the usage plan is more expensive at every volume including zero — the seat plan wins outright. If the unit rate is zero, the usage plan is flat and the two lines are parallel, so they never cross and whichever is lower stays lower. The calculator returns a blank crossover and says which case applies.
One subtlety trips people up. If your volume is inside the included allowance, the usage plan bills only the platform fee, and small changes in volume cost nothing at all. The crossover still sits above the allowance, so the comparison remains valid, but the marginal cost of a unit is zero until you exhaust the bundle — which is exactly when teams stop watching consumption and get surprised later.
Worked example: 40 seats at $25 against $300 plus $0.004 a unit
A vendor offers 40 seats at $25 a month, or a consumption plan with a $300 platform fee, 50,000 units included and $0.004 per unit above that. You expect 300,000 units in a typical month and 520,000 in your busiest.
- Seat plan. 40 × $25 = $1,000 a month, whatever happens.
- Billable units at expected volume. 300,000 − 50,000 = 250,000 units.
- Usage plan at expected volume. $300 + 250,000 × $0.004 = $300 + $1,000 = $1,300 a month.
- Difference. $1,300 − $1,000 = $300 a month, or $3,600 a year, in favour of the seat plan.
- Crossover. 50,000 + ($1,000 − $300) ÷ $0.004 = 50,000 + 175,000 = 225,000 units. Check it: $300 + 175,000 × $0.004 = $1,000, exactly the seat plan.
- Peak month. $300 + (520,000 − 50,000) × $0.004 = $300 + $1,880 = $2,180, which is $1,180 above the fixed seat plan in a single month.
- Effective unit price on seats. $1,000 ÷ 300,000 = $0.003333 per unit, against $0.004 on the consumption plan.
Step 7 is the cleanest way to express the result: at your expected volume you are already buying units at $0.003333 on the seat plan, and $0.003333 ÷ $0.004 − 1 = −16.7% against the consumption rate. You would have to fall to 225,000 units a month — a quarter below today — before consumption became cheaper, and the peak month exposes you to $1,180 of variance for the privilege.
How to choose once you have the crossover
Compare the crossover against your growth direction, not just your current position. If usage is growing and you are below the crossover, consumption is cheaper today and will stop being cheaper on a date you can estimate. If usage is falling and you are above it, the seat plan is protecting you now and will look expensive later. In both cases the contract term should be shorter than the time to cross.
Weigh the peak-month figure against how your budget works. A team with a fixed annual budget and no mechanism for overspend should treat variance as a real cost, and a fixed seat plan that is 10% more expensive on average may still be the right answer. A team that can pass consumption through to a customer or a cost centre should weigh it much less.
Look for the shape of the vendor's incentives too. Consumption plans usually include commitments, tiers and overage rates that change the slope, and a committed spend tier is simply a per-seat plan wearing different clothes — fixed money in exchange for a lower rate, with the same utilisation risk that seat contracts carry. Price any committed tier as a floor, not a discount.
Finally, check that seats and units measure the same work. A seat plan is priced on people and a usage plan on machine actions, so automation makes the seat plan cheaper per unit and the usage plan dearer. If you are about to automate a workflow that multiplies unit volume without adding people, the crossover you calculated today will move sharply against consumption.
Cost each way across a range of monthly volumes
| Monthly units | Seat plan | Usage plan | Usage $/unit | Seat $/unit |
|---|---|---|---|---|
| 50,000 | $1,000 | $300 | $0.006000 | $0.020000 |
| 100,000 | $1,000 | $500 | $0.005000 | $0.010000 |
| 150,000 | $1,000 | $700 | $0.004667 | $0.006667 |
| 200,000 | $1,000 | $900 | $0.004500 | $0.005000 |
| 225,000 | $1,000 | $1,000 | $0.004444 | $0.004444 |
| 300,000 | $1,000 | $1,300 | $0.004333 | $0.003333 |
| 400,000 | $1,000 | $1,700 | $0.004250 | $0.002500 |
| 520,000 | $1,000 | $2,180 | $0.004192 | $0.001923 |
The two unit-price columns meet at the 225,000-unit crossover, which is the arithmetic check on the whole table. Above it the seat plan is cheaper per unit and the gap widens; below it the usage plan is.
What to check in the contract before comparing
- What exactly is a unit. An API call, a record processed, a message sent and a credit are all called units by different vendors, and one action can consume several. Get the definition in writing before you model it.
- Whether the allowance resets or rolls over. A monthly allowance that expires unused is worth much less than one that pools across a year, particularly for seasonal workloads.
- What the overage rate is. Consumption above a committed tier is frequently priced above the standard rate, which bends the line upwards exactly where you are most exposed.
- Whether the platform fee includes support. A cheaper platform fee with paid support is not cheaper; fold support into the fixed term of both models.
- How the seat plan defines a seat. Named, concurrent and monthly-active definitions produce very different counts from the same team.
- Whether there is a floor or a ratchet. Some consumption contracts set the next period's minimum from the current period's peak, which converts one busy month into a permanent cost increase.
Key terms
- Two-part tariff
- A price with a fixed component and a variable component — here the platform fee plus the per-unit rate. It is the standard shape of consumption pricing.
- Crossover volume
- The monthly usage at which both plans cost the same. Below it the consumption plan is cheaper; above it the seat plan is.
- Effective unit price
- A plan's monthly cost divided by the units consumed. It is the only way to compare a seat quote and a consumption quote directly.
- Committed tier
- A discounted unit rate bought by promising a minimum spend. It reintroduces the utilisation risk that consumption pricing was supposed to remove.
Which model to prefer when the numbers are close
When the two plans land within a few percent of each other, choose on volatility and on who benefits from your own efficiency work. Consumption pricing rewards optimisation directly — cut redundant calls by a fifth and the bill falls by a fifth — while a seat plan captures none of that. If you expect to invest in efficiency, consumption pays you back for it.
Seat pricing wins where the team is stable, usage is spiky, and the budget is fixed. It also wins where usage is driven by automation you do not fully control, because the failure mode of consumption pricing is an unbounded bill from a bug. If you take a consumption plan, put a hard alert on daily spend and know whether the vendor offers a cap.
Whichever you choose, the decision has a shelf life. Recalculate at each renewal alongside the uplift terms, check your seat utilisation with the seat calculator so the seat side is not inflated by shelfware, and fold the whole thing into a total cost of ownership view if implementation and admin time differ between the two models — they often do, because consumption plans need monitoring that seat plans do not.