Consistency belongs in the work that is the same in every practice. It does not belong in the operatory. Finance, payroll, contracting, purchasing, and data security get stronger when they are built once and run by people who do nothing else. Clinical judgment does not work that way.
The operating principle is short: standardize the center, not the chair. Applying it is harder than saying it, because the line between center and chair moves depending on the function. Four categories make the line usable.
The test that decides where work belongs
Before any function moves out of the practice, ask whether the practice gets stronger. The question is not whether the group gets tidier, and not whether the reporting looks better at the top.
Stronger has visible signs. The doctor spends less time on work that does not require a doctor, and the team gets faster answers than it used to. Errors fall, and costs become visible instead of estimated. Hiring gets easier, and coverage exists when someone is out.
That test can be failed, which is what makes it a test. A shared function that produces reports nobody uses has failed it. So has an approval step that slows purchasing without lowering the price. If moving work to a platform does not make the practice stronger, the platform has not earned the work, and the work should go back.
Work that is stronger when it is built once
Some functions carry no clinical content at all. They are technical, they look much the same in every state, and they reward specialization. Running them separately in each practice means paying for the same expertise many times over, and getting a thinner version of it each time.
- Finance and accounting. One chart of accounts, one close calendar, one set of statements that can actually be compared.
- Payroll and benefits administration. Compliance work that gets more complicated as headcount grows, and more expensive when it goes wrong.
- Contracting. Payer agreements, vendor terms, equipment leases, real estate. Most owners negotiate these a handful of times in a career, and someone should be doing it every week.
- Purchasing. Supplies, lab, equipment. Price matters, and so does knowing the same materials will be on the shelf next month.
- Technology and data security. Backups, access control, patching, incident response, breach planning. A single practice cannot staff this properly, and the consequences of getting it wrong are not proportional to practice size.
- Reporting. Numbers defined the same way and produced on the same schedule, so a comparison means something.
This is what a shared operating platform is for. None of it touches a patient, and all of it improves with focus.
Standards shared because inconsistency creates risk
The second category has a different shape. The work stays in the practice and is done by the practice’s own team. The standard is common, because inconsistent execution creates exposure that everyone ends up carrying.
- Revenue cycle workflow. Verification, coding, claim submission, and follow-up on aging claims. Variation here shows up as denied claims and money left uncollected.
- Safety and infection control. There is no local version of this that is better for being local. Regulation sets the floor, and the standard should sit above it.
- Common definitions. What counts as a new patient. When a treatment plan is closed. What goes into production. Without shared definitions, every comparison across practices is fiction.
- Onboarding. A new hire should meet the same expectations in the first week, whichever office hired them.
Notice the shape. The standard is common and the execution is local. That is not the same as running an office from somewhere else. A practice that meets the standard its own way has met the standard.
Support that should be available rather than imposed
The third category is capability a practice can use or decline. Recruiting, mentorship, and continuing education belong here.
Recruiting is the clearest case. A shared function can source candidates, screen them, and handle the logistics that consume an owner’s time. It should not choose the associate or the hygienist. Fit is local knowledge, and the person who lives with the decision should make it.
Recruiting follows the same rule. A practice whose name people in town already trust should not be repainted to match a system template. That name belongs to the practice.
Mentorship and continuing education work when they are offered and stall when they are assigned. A doctor who wants a second read on a difficult case will go looking for one. A doctor scheduled into a program by someone else will attend it and forget it. The useful version of this support is a standing offer with a real person behind it, not a requirement with an attendance record.
The test for this category is simple. If a practice declines the support and no one else is harmed, the support is optional. That distinction is worth confirming line by line against what a support organization actually handles.
Work that stays in the practice
Clinical judgment, treatment planning, standard of care, patient relationships, and team culture all stay in the practice. They stay there permanently, not until a group reaches a certain size.
The reason is not sentiment. No committee has seen the patient. Nobody outside the operatory knows the medical history, the anxiety, the budget, or the years of visits behind this one. A treatment plan written by policy is a worse plan. Culture works the same way. It is built by who a doctor hires, keeps, and corrects, and it cannot be issued from an office somewhere else.
This protection lives in the documents rather than in the conversation. A promise about autonomy made in a meeting is a sentence, while a protected right is a clause you can point to later. In the terms of a doctor equity partnership, the questions worth tracing are where clinical authority sits and what can change it. Every partnership is governed by its final transaction documents.
The failure mode is real
Centralization goes wrong when it begins serving the center instead of the practice. It rarely announces itself, and it usually shows up first as friction.
- Requests take longer than they did before and pass through more people.
- Information flows up and does not come back down in a usable form.
- A standard gets applied to something where the variation had a good reason behind it.
- Vendors get consolidated in a way that simplifies the center’s work and raises the practice’s cost.
- There is no way to decline, and no way to appeal.
Be honest about the trade. Standardization costs you some control over how work gets done. That is a fair trade when the work gets done better, faster, or cheaper, and a bad one when it does not. The difference is measurable, so it should be measured. Measuring it means naming the expected outcome before the change, not judging it afterward from memory. A decision that fails should be reversible.
What to ask before work moves
- Which functions move to the platform, and which stay with me? Ask for the list in writing.
- Who decides when that list changes, and what say does the practice have?
- What is optional, and what happens if I decline it?
- How does a decision get reversed once it stops working?
- Where is clinical authority written down, and what could change it?
Vague answers to these questions are still answers. The sequence a partnership follows should make the boundaries clearer at each step, not less clear.
If you are weighing a partnership, sort your own functions into the four categories first. Then compare that list with what a group proposes to take. Where the two lists disagree, you have found the subject worth talking through, and that is a reasonable place to start a conversation.