Growing locations without growing your back office
In most dental groups, back-office headcount scales roughly linearly with locations. That relationship is not a law of nature — it is a consequence of how verification and claims work get distributed, and it is the first thing that breaks when a group crosses about ten locations.
The early-growth version works fine. Each office has someone who knows the local payer mix, keeps a mental model of which portals are reliable, and works the phone when they aren't. Add a location, add a person, and the system holds.
Then the arithmetic turns. Ten locations means ten sets of payer logins, ten interpretations of the same plan rules, and ten people who each independently learned that a particular carrier's portal understates waiting periods. Institutional knowledge lives in individuals, so it leaves when they do.
The three costs nobody budgets for
Variance
The same plan gets verified to different standards at different offices. One location catches a downgrade provision; another doesn't. The result is inconsistent estimates, inconsistent case acceptance, and a denial rate that varies by site for reasons no report explains.
Knowledge that doesn't compound
When a specialist learns a payer quirk, that knowledge stays local. Across a group, the same lesson gets learned separately many times over — and re-learned after every departure. Turnover in back-office roles is high enough that this is a recurring cost, not a one-time one.
The coordination tax
Centralizing into a shared services team helps with variance, but introduces queueing, handoffs, and a management layer. Many groups discover that centralizing moved the cost rather than removing it.
What actually breaks the linear relationship
Only two things decouple back-office headcount from location count.
First: the repetitive portion has to be genuinely automated, not merely assisted. A tool that drafts a verification for a human to review still needs a human per unit of volume. The staffing curve stays linear; the slope just gets slightly better. Automation only changes the shape when the work completes without a person in the path for the common case.
Second: payer knowledge has to live in the platform, not in people. When a payer changes a portal or a plan rule, that should be learned once and applied everywhere — every location, immediately. This is the compounding advantage that headcount can never reproduce, because a person's learning is bounded by the cases they personally see.
What this looks like in practice
A group running verification and claims on an agentic platform adds a location by adding the location's payers and PMS connection — not by adding a verification specialist. The remaining human work concentrates in judgment: complex plans, appeals, and the exceptions that genuinely need a person. That work grows sublinearly, because edge cases scale with plan diversity rather than with patient volume.
DentalRobot's platform automates up to 90% of repetitive back-office work and completes up to 95% of insurance verification end-to-end, with dental RCM specialists handling the remainder. Every verification and claim processed since 2018 feeds back into the system, so payer quirks become institutional memory that no departure erases.
The reference point that matters more than any percentage: Phil Toh at The Smilist, a 100+ location DSO, cut back-office verification labor by more than 80% — while continuing to add locations.
The question to ask before your next hire
When the next location opens and the plan calls for another verification specialist, the useful question is not whether that person is affordable. It is whether the role will still need to exist at location thirty. If the answer is yes, the cost is not one hire — it is the slope of a line you will be paying for as long as you keep growing.