Expensify Went Flat, and That Is the Honest Version of Usage Pricing

Most SaaS pricing punishes you for the accounts you are not using. You provision a seat, the seat sits idle, and the invoice charges for it anyway. Expensify does something different, and it is worth a CFO's attention not because expense software is exciting but because the billing model is the one more vendors should copy. As of July 2026, the Collect plan is $5 per member per month, flat, and it bills on active members: people who submitted at least one expense that month. Our catalog carries Expensify at $5 a member, which lines up with Collect. The interesting part is the word active.

What active-member billing actually means

Under active-member billing, a member who did not submit an expense this month does not appear on this month's bill. Not at a discount. At zero. You are charged for the people who used the tool, measured by the thing the tool exists to do (submit an expense), and nothing for the accounts that sat quiet.

Compare that to a standard per-seat tool. There, the unit you pay for is the provisioned seat: an account that exists in your identity provider. Whether the person behind it logged in this month is the vendor's good fortune, not your credit. The gap between provisioned seats and active users is exactly where SaaS waste accumulates, and on most tools it is invisible until you go pull the login data yourself. We wrote about that gap in your invoice hides three seat counts: the dormant licenses between provisioned and active are the seats you pay for and nobody opens.

Active-member billing closes that gap by construction. The dormant account cannot bill you, because billing is triggered by an action, not by the existence of an account. That is what usage pricing is supposed to be: you pay for what you consumed. Too much of what gets sold as "usage-based" is really a consumption meter bolted onto a seat minimum, where the seat still bills whether you use it or not. Expensify's Collect plan drops the seat and keeps the action.

Worked example: 15,000 employees, two billing models

Take an enterprise with 15,000 employees. Finance rolled Expensify out broadly, so 6,000 people have member accounts. In a normal month, 3,800 of them actually submit an expense. The other 2,200 have accounts they are not using this month: field staff between trips, employees who expense a few times a year, people who were provisioned in an onboarding batch and rarely file.

Under Expensify's active-member billing at $5, you pay for the 3,800 who submitted:

• 3,800 active members x $5 x 12 = $228,000 a year

Now imagine the same 6,000 accounts on a conventional per-seat tool at the same $5, where the provisioned account bills regardless of activity:

• 6,000 provisioned seats x $5 x 12 = $360,000 a year

The difference is the 2,200 idle members, which cost you nothing under active billing and $132,000 a year under per-seat billing. Same tool, same $5 rate, same company. The billing trigger alone is worth $132,000, and it is worth that every year without a single renegotiation, because the model never charges you for the idle accounts in the first place.

That is the honest version of the point we make constantly on the per-seat side: the waste is not the rate, it is the unit you are billed on. Change the unit from "account exists" to "account did something" and a large slice of the waste stops existing.

The Control tier, and where the commitment sneaks back

Expensify's higher tier tells the other half of the story. Control is $9 per member per month if you put spend on the Expensify Card and sign an annual commitment, or $18 per member per month without the card. The card funds the discount: Expensify makes interchange on the card spend, so it hands back half the per-member price to get it. That is a legitimate trade, and for a finance team already consolidating card and expense it can pencil well.

But read the annual commitment carefully, because it is where per-seat risk creeps back into a usage model. A month-to-month active-member plan bills only your submitters. An annual commitment can lock a member count for the term, which reintroduces exactly the provisioned-but-idle exposure that the active-member model removed. The $9 headline is real, and so is the string attached to it. If your active-member count swings seasonally, an annual commit sized to your peak month bills you the peak all year. Price the commitment against your trough, not your peak, or you have traded a clean usage model for a seat minimum wearing a discount.

What to check in your own expense-tool contract

Whether you run Expensify or a competitor, three questions decide whether you are on honest usage pricing or a seat count in disguise.

Which definition of member am I billed on? This is the whole game. "Active member" defined as submitted-an-expense is a usage trigger. "Member" defined as provisioned-account is a seat. Vendors use the same word for both. Find the definition in the order form, not the marketing page, and confirm what counts.

Am I on a monthly active model or an annual commitment? Month-to-month active billing self-corrects when usage drops. An annual member commitment does not. If you signed a commitment, check whether it was sized to peak-season activity and whether you are carrying that peak through your quiet months.

Am I on the current plan? Expensify's flat active-member Collect terms apply to newer signups. Older contracts may sit on prior structures with different triggers. If your paperwork predates the current plan, the definition of member you are billed on may not be the one on today's pricing page, and a plan migration can be worth more than a rate negotiation.

The broader lesson travels well past expense software. The vendors moving to outcome and consumption billing are, at their best, doing what Expensify does here: charging for the event instead of the seat. At their worst they are metering usage on top of a seat floor that never went away. We track that split in the shift to outcome-based pricing, in the renegotiation levers for usage-priced tools, and in where per-resolution agents cross over. When a vendor tells you the pricing is usage-based, the follow-up is always the same: usage of what, and does anything bill when I use nothing? If the answer is that idle accounts bill zero, you are looking at the honest version. If idle accounts still bill, you are looking at a seat with better marketing.