Why we bill per minute instead of per seat

We charge nothing per month. The meters run at $0.01 per video minute, $0.008 to $0.012 per minute of AI notes depending on whether the session is on video or in person, and $0.03 per SMS. Email is free. Card processing passes through at cost, 2.9% plus 30 cents, with no markup from us. Staff logins, appointments, and intake forms are unmetered, and the whole thing is capped at $69 per clinician per month. New practices start with a $100 credit and no card on file.

That is unusual enough in this market that it needs an explanation rather than a tagline. Here is ours, including the part where it works against us.

What a per-seat invoice actually measures

A per-seat subscription sends the same invoice in August, when a therapist is on vacation and running six sessions, as it does in October, when they are at capacity with a waitlist. The software did nothing different in either month. The therapist did.

For a large group practice that averages out and nobody notices. For a solo clinician it is the difference between a cost that tracks the practice and a cost that sits on top of it. Behavioral health is seasonal, caseloads move, people take leave, and a new practice has months where the schedule is genuinely thin. Charging a flat license fee through all of that means the months when money is tightest are exactly the months the software takes the largest share.

The starting cost is where this lands hardest. At published per-seat prices, a year of practice management runs $588 to $948 before a clinician has seen a single client: SimplePractice lists tiers from $49 to $79 a month, paid for twelve months whether the schedule fills or not. That is a real number to a therapist opening a practice with a laptop and a sublet office hour.

What metering does to our incentives

When revenue comes from seats, the growth lever is selling more seats and moving practices up tiers. Features get sorted by which plan they belong to, and a chunk of engineering time goes into the machinery that enforces the sorting.

Metering points us somewhere narrower. We make money when sessions happen. A no-show costs us the same minutes it costs the clinician, which is why appointment reminders, easy rescheduling, and a video room that connects on the first try are not upsells here. They are how we get paid. If a practice stalls, our revenue from it goes to zero, and we find out fast rather than collecting a license fee from an account that quietly stopped using the product.

It also removes the argument about who gets a login. Billing per seat makes a practice ration accounts, so the office manager ends up sharing credentials with a virtual assistant and the audit trail becomes fiction. We would rather everyone have their own account with the right permissions, so we do not charge for accounts.

Why there is a cap

Usage pricing has an obvious failure mode: a clinician doing mental arithmetic during a session about whether to let it run long. That is a terrible thing to introduce into a therapy hour, and any pricing model that produces it deserves to lose.

So we capped it at $69 per clinician per month. Past that point the month is free, however many sessions run. Work the numbers for a full caseload and the cap binds quickly: 80 telehealth sessions of 50 minutes is 4,000 video minutes at $40, plus $32 of AI notes, which comes to $72 and bills as $69. A 30-session month costs about $27. The busy clinician pays the ceiling; the one rebuilding after a slow quarter pays about 40 percent of it.

Which means the busiest clinician on the platform will be our least profitable per hour. We designed it that way, and it is the piece of the model most likely to get argued about internally as we grow.

What this costs us

Predictable recurring revenue, mostly. Per-seat is easier to forecast in every direction: you know December's revenue in November, churn is a clean signal, and the whole apparatus of SaaS planning assumes it. Metered revenue moves with the season, with holidays, with a flu week that empties three schedules. We will have quarters where revenue falls because our customers took time off, and we will have to be fine with that.

Investors prefer the other model, and they are not wrong to. A subscription business with 90% gross margin and predictable expansion is simply easier to underwrite than one whose top line tracks how many therapy hours happened in America last month.

The honest version: we have not proven this works at scale. Valence is early, and the model is a bet that alignment with small practices beats the forecasting comfort of per-seat. If metered revenue turns out too lumpy to run a company on, we will have to change something, and we would rather write that here than pretend to a certainty we do not have. What we can commit to is that any change gets announced in advance and does not arrive as a surprise line on an invoice.

How to check the math on your own practice

Do not take the argument on faith. Put your real numbers into the pricing calculator, which runs the same rates the backend meters against, and compare the result to whatever you are paying now. If your practice is large and busy enough that the cap binds every month, the comparison gets closer, and we would rather you see that in the calculator than discover it later.

If you are still deciding what to run a new practice on, we also wrote about choosing an EHR for a new private practice, which covers the parts of that decision that have nothing to do with pricing.