FMG schedule management: which tool to choose.

What separates an FMG schedule from a shift schedule

In a store or a restaurant, scheduling means assigning hours to hourly employees while respecting labour standards and feeding payroll. Nearly every scheduling product sold in Quebec is built on that model, and most do it well.

An FMG works differently. Most physicians are self-employed or incorporated, and they bill RAMQ rather than the clinic. Money does follow the schedule, but through billing rather than an hourly payroll run, with no time clock, no overtime and no labour cost to optimize. What the schedule produces first is coverage. Someone has to hold the Tuesday walk-in block from 1 p.m. to 4 p.m., someone has to take the weekend call, and those decisions have to look fair to a team whose manager directs no one in the hierarchical sense.

Three everyday FMG objects simply do not exist in a shift tool. The walk-in block, which must be covered no matter who takes it and which weighs on the fairness balance. The on-call shift, whose value depends on when it falls, since a Friday evening in December is not a Wednesday morning in May. The shared office, because an FMG almost always has more professionals than rooms, and a schedule with no room is an imaginary schedule.

On top of that sit the multidisciplinary team, the affiliated sites to coordinate with the main site, and the ministerial framework. Quebec's FMG funding and professional support program, in the version in force since 1 April 2026, ties funding to accessibility obligations that translate into blocks spread across all opening hours, evenings and weekends included. The same program sets a floor of 68 hours of services per week spread over seven days, with a family physician or an IPSPL present at all times, and an accès-réseau FMG adds 28 to 84 hours on top depending on its designation level.

The six families of tools

Product names change every couple of years, families barely move. Placing a vendor in the right box does half the evaluation.

Family	Examples	What it does well	Where it breaks in an FMG
Spreadsheet	Excel, Google Sheets	Free, flexible, no training, no one to convince	No optimization, fairness never calculated, dependent on whoever masters the file
Shared calendar and poll	Outlook, Google Calendar, Teams, Doodle	Broadcasting the schedule, collecting availability	Displays decisions taken elsewhere, detects neither conflicts nor imbalance
Workforce scheduling tool	Agendrix, Voilà!, Emprez, Skello	Time clock, availability, swaps, payroll prep, labour standards	Salaried employee model, no weighted call, no offices, no walk-in logic
Broad health platform	Petal and other institutional suites	Covering several departments, fitting a hospital ecosystem	Quote-based pricing, institutional procurement, project team required
Specialized medical team scheduling tool	Synchro	Blocks, call, offices, fairness and transfers in one model	Replaces neither the EMR nor an hourly payroll
Tool built in-house	Home-built application, written with an AI assistant or contracted to a developer	Fits your rules exactly, no licence to pay	The clinic becomes a software publisher: maintenance, security patches, bus factor and Law 25 land on it

Workforce scheduling tools are good at their job. For hourly staff, receptionists and administrative assistants first, an Agendrix or a Voilà! handles time clocks, replacements and payroll export better than a medical tool would. The trap appears when physicians get pushed into the same model. Call becomes an ordinary shift, fairness shrinks to a headcount, and the manager reopens the spreadsheet to settle the real cases.

Broad platforms are not oversized by accident, they answer to an organization with an IT department, a tendering process and several departments to cover. A 25-professional FMG that runs its own schedules has none of that structure, and the gap shows at onboarding. Our Synchro and Petal comparison covers that split.

The sixth family is the most recent one and the easiest to underestimate. An AI assistant now produces a calendar application with a database and user accounts over a weekend, which makes the question "why pay a subscription" a serious one in clinics where it was not asked two years ago. The visible part gets built fast, and that is what misleads. What takes months is how the engine behaves when no assignment satisfies every constraint, which is the normal case, and everything that happens after publication: withdrawals, replacements, transfers, notifications. The clinic also becomes the publisher of the software, with security patches, the bus factor and Law 25 obligations that no longer have a vendor to carry them. Our Synchro and an in-house tool comparison prices the three paths line by line, including the cases where building is still the right call.

What about the EMR?

Ofys, Omnimed, Medesync and Kinlogix manage the patient record and appointment booking. They consume your schedule, they do not produce it. Blocks exist in the EMR because someone decided elsewhere who was covering them, and that elsewhere is exactly what you are trying to equip.

Eight criteria, in the order that matters

1. Does the tool build the schedule, or display it?

A shared calendar shows decisions once they are made. An assignment engine makes them, from each person's constraints, the blocks to cover and the distribution rules, and the difference is measured in management hours per month. In a demo, have a full schedule generated in front of you and watch what is left to fix by hand.

2. Is fairness calculated, or merely counted?

"Everyone has five call shifts" settles no argument, because five weekday shifts and five weekend shifts are not worth the same. Look for explicit weighting (evenings, weekends, statutory holidays, rotation across the year) and a view physicians open themselves. Invisible fairness gets contested, visible fairness gets discussed once and then settled.

3. The place of rooms in the model

If the tool ignores offices, the manager keeps a room plan on the side and redoes the arbitration every Monday morning. Check that a room is a resource in its own right, and that the tool refuses to put two physicians in the same office at the same hour.

4. How much effort does the tool ask of physicians?

This is the criterion that decides adoption, and it is routinely underestimated. A physician will open the app a few minutes a month, never more. If declaring "no call on Tuesdays" takes a forty-field form, the email to the manager wins again and you will have paid for a tool only the administration uses. Look at how constraints get declared, what the phone view looks like, and how many taps separate a physician from the answer to "when am I working next week".

5. What happens to the schedule when someone drops out?

A published schedule rarely survives intact, and the management cost hides mostly in last-minute replacements. A good solution lets the physician offer the shift to eligible colleagues, notifies automatically, and records who accepted and when. A poor one sends you back to the phone chain, with an official schedule that lies. The transfer board is the mechanism to watch closely.

6. What becomes of your personal information

Scheduling software holds no patient record, but a physician's availability, absences and preferences are still personal information under Law 25. Ask where the data is hosted, what is logged, and what is left of it thirty days after a cancellation. A vendor who hesitates has already answered.

7. French, terminology and support in your time zone

An interface translated from English gives itself away quickly. Gardes become "shifts", plages become "slots", and nobody on the team uses those words. Check that Quebec vocabulary is there (GMF, GMF-U, sans rendez-vous, garde, AMP, PREM) and that support answers during your opening hours, not the West Coast's.

8. Price, trial, and time to the first published schedule

A published price lets you compare without opening a sales conversation. A quote-based price is not a fault in itself, but it changes the nature of the evaluation. Either way, ask who counts as a user (the part-time physician, the receptionist, the resident), whether setup fees exist, what happens mid-year when the team grows, and how long it takes from signature to the first schedule actually published.

Eight questions to ask in a demo

Criterion	The question to ask
Assignment engine	Can you generate a full schedule in front of me, with my constraints?
Fairness	How do you weight a weekend call shift against a weekday one?
Rooms	What happens if I put two physicians in the same office at the same time?
Adoption	How many taps for a physician to declare a constraint from a phone?
Replacements	Show me a call transfer end to end, notifications included.
Law 25	Where is the data hosted, and what remains thirty days after cancellation?
Language and support	What are your support hours, and in which language do you answer?
Price and delay	Who counts as a user, and when will I publish my first schedule?

How to test before you sign

A demo shows the product at its best, which is normal. A well-run trial shows your FMG as it really is. Four to six weeks is enough, provided you pick a representative period rather than a quiet month.

1. Rebuild a schedule you have already published. You know its right answer and its compromises. Compare the result, not the promise.
2. Time it. Build time, back-and-forths with the team, corrections after publication. Those three numbers compare across tools, impressions do not.
3. Put two skeptical physicians in the room, not the two most enthusiastic. If they find their schedule in under a minute and declare a constraint unaided, adoption will follow.
4. Trigger an incident. Cancel a call shift during the trial and follow the replacement through. That is the scene that replays every week.
5. Get your data out before deciding. Export the schedule to CSV and iCal during the trial. A tool you cannot leave cleanly is a risk, whatever its qualities.

The mistakes we see most often

The most common one is choosing on the length of the feature list. A tool that ticks forty boxes but that no physician opens costs more than the spreadsheet it was meant to replace.

Next comes the decision taken without the physicians, then announced to them. Weekends and evenings are sensitive ground, and an imposed decision is paid back in passive resistance for six months.

Testing in July flatters every tool. Take a period with vacations to cover, a statutory holiday and at least one last-minute drop-out instead.

The front desk gets forgotten just as easily. Administrative staff live inside the schedule as much as physicians do, and a solution that does not speak to them creates a third parallel system.

Finally, a published price is rarely read as far as the definition of a user. A $3 per-person gap becomes significant once that definition counts residents, locums and read-only accounts.

Where Synchro sits

Synchro is a tool specialized in team scheduling: walk-in blocks, on-call shifts, shared offices, constraints each physician writes in plain French, transfers between colleagues. It holds no patient data, does not replace your EMR and does not run payroll. The price is published, $10 CAD per user per month billed annually and $15 CAD month-to-month, first month free, and configuring a 30-user team takes about 45 minutes.

A few limits, while we are being clear. The company was founded in 2025 and the product matured in a pilot FMG in Saint-Hyacinthe. If your organization requires a multi-department platform, or has specific hosting requirements, say so early in the conversation. Our security and compliance page answers the technical half.

Read next

• FMG scheduling software: the full guide
• Why Excel is not enough for FMG scheduling
• Synchro and Petal: factual comparison
• Synchro and Agendrix: medical tool or HR tool
• Synchro and Doodle: poll or schedule
• Synchro and an in-house tool: build or buy
• Pricing