A clinic pays for five systems. A new platform appears to cover work performed in three of them. The tempting calculation subtracts three subscriptions from one quote and calls the difference ROI.
That arithmetic is premature. A contract is not retired when a feature appears in a demo. It is retired when the replacement completes the real workflow, the practice can retrieve its records, staff know the exception path, and the old vendor can be switched off without losing required work.
Use a Practice Copilot ROI case to decide what the clinic can retire, what it must retain, and what evidence is still missing. Keep the spreadsheet small enough that an owner can challenge every number.
Start with a three-column software ledger
List every tool touched by one patient journey, from booking through the signed record and approved follow-up. Put each tool in one of three columns.
| Decision | Meaning | Evidence needed |
|---|---|---|
| Retire | The proposed workflow replaces the job and the contract can end | Completed acceptance test, export plan, owner approval, cancellation date |
| Retain | The clinic still needs the system | Named job, current cost, interface or manual handoff |
| Overlap | The answer depends on a test or contract term | Open question, test date, decision owner |
Do not use “included” as a fourth decision. Included describes the quote. Retire describes an operational result.
Patient Square currently publishes three US tiers. The monthly prices shown in September 2026 are $99 for Assist, $149 for Copilot, and $249 for Autopilot. Annual billing is shown as $79, $119, and $199 per month respectively. The current pricing page is the source to check before a purchase decision because prices and plan scope can change.
The same page lists documentation and scheduling in Assist, Patient Square EHR and WhatsApp in Copilot, and named receptionist and follow-up modules in Autopilot. Those labels still need to be translated into the clinic’s work.
Calculate a conservative first-year case
Use four lines, not an elaborate forecast.
Confirmed annual savings from retired tools $________
Confirmed staff cost removed from named handoffs $________
New annual subscription $________
One-time implementation and transition cost $________
Conservative first-year value = savings - subscription - transitionOnly the first two lines require judgment. A software saving belongs in the model when the cancellation date is known and the clinic has handled export, retention, and closeout. Staff savings belong there when a task goes away, not when it merely moves to another screen.
Keep possible upside below the line. If shorter documentation work might create more visit capacity, show it as a scenario. Do not convert every saved minute into revenue unless the schedule, demand, staffing, and completed-visit data support that conversion.
Follow one visit before pricing the whole platform
AHRQ describes workflow mapping as a way to make a process visible, clarify who performs each step, and find opportunities for redesign. Its workflow mapping guide recommends mapping the current process before designing the future one.
Pick a common visit with a slightly awkward ending. For example:
- A patient books and later asks to change the time.
- Staff confirm identity and update the appointment.
- The clinician conducts the encounter and reviews a drafted note.
- The clinician signs the final record.
- Staff send an approved follow-up message.
- The patient replies after the office closes.
Mark the system, person, input, output, and exception owner at each step. Run the proposed setup with synthetic data. Then disable one expected handoff or send the reply to the wrong queue on purpose. The business case should include the labor and fallback that remain after that test.
Give each planned retirement a proof card
A one-line assumption is too easy to forget. Use a proof card for every system marked Retire.
Current tool and contract end: ______________________________
Job it performs today: ______________________________________
Replacement workflow tested on: _____________________________
Routine case result: ________________________________________
Failure case result and owner: ______________________________
Data export and retention checked by: _______________________
Retirement approved by: ____________________________________If the card has blanks, keep the cost in the overlap column. This makes the ROI case less exciting and more useful.
Treat the record decision as its own gate
Patient Square describes two Practice Copilot deployment paths: use the Patient Square EHR or work alongside an existing EHR. That choice changes migration work, duplicate entry, export needs, and the number of systems the clinic can plausibly retire. The Practice Copilot page is explicit about the two paths; it does not establish an interface with every named outside EHR.
Draw a box around the system that will hold the completed, signed record. Write down how scheduling, draft notes, messages, and approved follow-ups enter that box. The ASTP/ONC SAFER Guides provide practical safety material for EHR configuration, communication, patient identification, contingency planning, and other areas worth testing during a change.
No retirement credit belongs in the model while the authoritative record or fallback route is undecided.
Present three outcomes to the owner
The final packet can fit on one page.
- Proceed: tested retirements cover the new cost and the owners accept the transition work.
- Pilot: the workflow looks viable, but important costs remain in the overlap column.
- Stop: the clinic would add a platform while keeping the same systems and staff handoffs.
Attach the software ledger, the conservative calculation, and one proof card per proposed retirement. That is enough for a real decision. A large feature matrix is not.
The strongest ROI case may conclude that the clinic should buy a smaller plan, keep its present EHR, or wait until a contract expires. It may also show that several subscriptions and handoffs can safely disappear. Either result is better than counting a demo as a saving.