At a single location, verification is usually a person rather than a process. Someone has been doing it for six years. She knows which payer answers fastest before ten, which portal logs you out mid-session, and which plan quietly downgrades molar crowns. None of it is written down, and it works, right up until she takes a week off.
Add locations and that arrangement does not scale, it multiplies. Ten offices do not share one process with ten sets of hands. They have ten processes that grew independently, each shaped by whoever set it up and whichever payers dominate that zip code.
This is not a niche problem anymore. ADA Health Policy Institute data from 2024 found more than one in four dentists within ten years of graduation were affiliated with a dental service organization, a rate well above older cohorts. Consolidation is arriving through the front door of the profession.
The symptom is data you cannot compare
The failure mode multi-site groups notice first is not a missed verification. It is a report that cannot be trusted.
A regional dashboard says 94% of tomorrow’s patients are verified. At one office that means eligibility was confirmed. At another it means a full benefit breakdown is documented, including frequency history and downgrades. Those are not the same achievement, and averaging them produces a number that describes nothing. Leadership makes decisions on it anyway, because it is the only number available.
Five things that drift, in roughly this order
Definition of done. Without a written field list, every office decides for itself when a patient counts as verified, and the definition tends to settle wherever the local workload forces it.
Payer mix. The hard carriers in one market are not the hard carriers in the next. A protocol built around the three payers that dominate one location is close to useless forty miles away, which is why locally-grown processes resist being copied.
Where the information lives. One office types the breakdown into the appointment note. Another uses the insurance tab. A third keeps a shared spreadsheet because the PMS field is too short. Every one of those choices is reasonable in isolation, and together they make the data unqueryable.
Escalation. How long does someone stay on hold before giving up, and what happens next? At one office it is forty minutes and a second attempt. At another it is ten minutes and a provisional estimate. Nobody wrote either rule down.
Coverage during absence. A single office absorbs a vacation informally. Ten offices with ten informal arrangements produce a predictable pattern of gaps that only shows up in the denial report a month later.
Write the specification before centralizing anything
The instinct is to fix this by consolidating verification into a central team or an outside vendor. That is usually the right destination, and it fails when it happens first.
Centralizing an undefined process does not standardize it. It relocates the inconsistency somewhere less visible, and now the offices cannot see it happening either.
What has to exist first is a specification, and it can be short. One list of fields that constitutes a complete verification. One location in the practice management system where each field lives, identical across sites. One definition of verified, phrased so that two people reading the same chart reach the same conclusion. One escalation rule with a time limit. One metric, measured the same way everywhere.
With that in place, centralization becomes an operational decision rather than a leap of faith, and it becomes possible to tell whether it worked.
Acquisitions make the case sharper. A practice you buy arrives with a verification process already running, built around its own payers, its own software habits, and one person’s memory. The temptation is to leave it alone during the transition, because it appears to be working and there are louder problems. Six months later it is a permanent exception, and the group has eleven processes instead of ten. The specification is what gives an incoming office something concrete to convert to.
The test
Pull the same patient plan at two locations and compare the documentation side by side. If the two records contain different fields, in different places, with different levels of detail, the problem is not staffing or effort. It is that nobody has said what the job is.
That is a document, not a system. It costs an afternoon to write, and everything downstream depends on it existing.
Source
ADA Health Policy Institute, 2024 data — more than 1 in 4 dentists within 10 years of graduation affiliated with a DSO
https://adanews.ada.org/new-dentist/2025/november/hpi-more-new-dentists-affiliated-with-dsos
