Almost every EMR quote you receive will use one of four pricing structures: per-provider subscription, per-encounter, a percentage of collections, or a perpetual license with annual maintenance. The structure matters more than the headline number, because the structure decides what happens to your bill in the years when you add a part-time provider, lose one mid-term, open a second location, or have an unusually good year. Compare structures against your own projections, not first-year totals against each other.
The four models you will actually see
| Model | How you are billed | Tends to suit | The thing that bites |
|---|---|---|---|
| Per-provider subscription | Recurring fee per licensed user, often tiered by role | Practices with stable, predictable headcount | How "user" is defined. Locums, per-diem, part-time, billing staff, and read-only accounts are all negotiable definitions, and vendors define them in their favor by default. |
| Per-encounter or per-visit | Fee per visit, encounter, or claim | Low-volume, seasonal, or fast-growing practices | It taxes growth. Model it at your projected year-three volume, not today's. |
| Percentage of collections | A share of what the practice collects, usually bundled with billing services | Practices outsourcing revenue cycle alongside the software | Your software cost rises with revenue regardless of the software's effort, and the software price is invisible, so you cannot compare it to anything. |
| Perpetual license plus maintenance | Large upfront license, then recurring maintenance | Practices with IT capacity and a preference for capital spend | Maintenance escalators, hardware refresh cycles, and whether maintenance is calculated against list price or your discounted price. |
How each model behaves when you change
Run every quote through the four events most likely to actually happen to you over a three-year term.
- You add a provider. Per-provider is immediate and linear. Per-encounter only bills you when the new provider produces. Percentage of collections is the same story. Perpetual license usually means buying seats.
- You lose a provider mid-term. This is the one buyers never check. Most subscriptions do not let you reduce seat count mid-term at all, and many do not let you reduce it at renewal without notice. Ask specifically for a true-down right and a notice window you can actually meet.
- Volume spikes. Per-encounter and percentage models grow with you, which is either fair or painful depending on your margin. Per-provider is flat, which is the point of it.
- You open a second location. Ask before you sign. Additional-location fees are common, rarely volunteered, and awkward to negotiate once you are locked in and visibly expanding.
What is almost never in the base price
- Implementation and configuration
- Data migration, often priced per record, per chart, or per hour
- Interfaces: lab, imaging, practice management, clearinghouse. Usually a one-time fee per interface plus a recurring maintenance fee per interface
- Training beyond a stated number of hours
- E-prescribing, and separately the controlled-substances capability and its identity-proofing or token costs
- Patient portal, appointment reminders, and text messaging
- API access to your own data, and electronic export of it
- Premium or after-hours support tiers
- A test or sandbox environment
- Custom report building
- Data extraction at termination, in a usable format
Normalizing competing quotes
Vendors do not present quotes in comparable shapes, and that is not an accident. Force every one of them into the same worksheet, over the same window, using your numbers.
| Line item | Vendor A | Vendor B | Notes |
|---|---|---|---|
| One-time: implementation and configuration | Get it fixed-fee or capped, not "estimated" | ||
| One-time: data migration | How many years of history? Discrete data or PDFs? | ||
| One-time: interfaces | Multiply by the number you actually need | ||
| One-time: training | Hours included, and the rate beyond them | ||
| Recurring: subscription over 36 months | Use your year-two and year-three headcount or volume, not year one | ||
| Recurring: interface maintenance | |||
| Recurring: support tier | Price the tier that matches your operating hours | ||
| Recurring: e-prescribing and add-on modules | |||
| Escalation | Apply each vendor's contractual cap. No cap is a finding, not a blank. | ||
| Exit: data extraction | If they will not quote it now, they will quote it when you have no leverage. | ||
| 36-month total |
Three rules make this worth doing. Use a 36-month window at minimum, because promotional first-year pricing is designed to survive a 12-month comparison. Use your own projected provider count and visit volume for years two and three rather than today's. And price the exit before you price the software, because the cost of leaving is the only line item the vendor has no incentive to make attractive.
Questions that flush out hidden cost
- Exactly who counts as a billable user? Does a provider working two days a week count as one?
- Can we reduce seats at renewal? Mid-term? What notice is required, and in what form?
- Is the annual increase capped in writing, for the initial term and for renewal terms?
- What is the fee per interface, one-time and recurring, and which interfaces do we need for the workflows we just demoed?
- What does it cost to get all of our data out, in what format, and how many days does it take?
- Is API access to our own data included?
- Are certified-module upgrades included in maintenance, or repriced as new versions?
- What is the fee to add a second location?
- What is the termination fee, and how is it calculated?
Red flags in a quote
- Year-one-only pricing. A promotional rate with no stated year-two rate is not a quote, it is an invitation.
- Auto-renewal with a long notice window and no price cap. The combination is the point. Missing the window renews you at a price nobody agreed to.
- Bundled percentage-of-collections with no itemized software price. You cannot compare it, you cannot benchmark it, and you cannot leave the billing service without also leaving the EMR.
- "Data migration: TBD." This is the single most common source of budget overrun in an EMR purchase.
- A termination fee that exceeds the remaining subscription value.
- No stated cost for data extraction at exit.
The takeaway
The pricing model is a bet on your own future. Per-provider bets on stable headcount, per-encounter and percentage models bet on modest growth, and a perpetual license bets on your capital and your IT capacity. There is no universally cheapest structure, only the one whose failure mode you can live with. Force every quote into the same 36-month worksheet with the same assumptions, insist on a written escalation cap, and get the exit priced while the vendor still wants your signature.
Common questions
Which EMR pricing model is cheapest?
None of them universally. Per-provider tends to favor stable, high-volume practices; per-encounter tends to favor low-volume or seasonal ones. The only way to know is to model each quote over 36 months using your own projected headcount and visit volume.
Should we bundle billing services with the EMR?
Bundling can simplify operations, but insist on an itemized software price even inside a bundle. Without it you cannot compare vendors, and you cannot change billing services later without also changing your EMR.
Are EMR prices negotiable?
Usually more so on implementation fees, training hours, interface counts, and renewal price caps than on the headline per-user rate. Renewal caps in particular are commonly negotiable and are worth more over a full term than a small discount on year one.
What should we budget for beyond the subscription?
Implementation and configuration, data migration, each interface, training, a temporary drop in productivity around go-live, and the cost of extracting your data if you ever leave.