How Much VA Support Does a Dental Practice Need?

Practices systematically underestimate because they count only the work they can see. The invisible work (recall not worked, claims not followed, treatment not chased) is precisely the work the support is meant to do, and it does not appear in anyone’s current hours. The intuition that the practice needs “about 20 hours per week” is usually an estimate of visible overflow, not of the full body of work. The bottom-up method below corrects for this by working from volume drivers instead of gut estimates.

Step 1: Estimate current visible administrative hours

Time-log all administrative staff and the owner for two weeks. The two-week window captures variation better than a single week and smooths week-specific anomalies. The log covers every administrative task performed: front desk, insurance, recall, communications, reporting, management. The output is a realistic estimate of current visible administrative load.

The first thing most owners discover at Step 1: the owner is contributing more administrative hours per week than they realized. Some of that is unavoidable and belongs with the owner. Some of it is the delegation gap the VA is meant to address. The two-week log makes that gap visible for the first time in most practices.

Step 2: Estimate the invisible work

This is the work that is currently not being done. Use volume drivers to estimate what the work would take if it were being done at an appropriate standard.

The time figures in the table below are analyst estimates and must be replaced with practice-specific measured data before acting on them. They are illustrative structure, not benchmarks. Do not treat them as industry standards or targets.

Workflow	Volume driver	Rough time per unit
Benefits verification	Appointments per week	10 to 15 minutes each
Claim status follow-up	Open claims	5 to 10 minutes each
Denial work	Denials per month	20 to 30 minutes each
Recall outreach	Due patients per month, multiplied by touch attempts	3 to 5 minutes per touch
Unscheduled treatment follow-up	Open treatment plans, multiplied by touch attempts	5 minutes per touch
Lapsed patient reactivation	Lapsed patients in a campaign period	3 minutes per touch
Huddle preparation	Daily	45 to 60 minutes per day

To use the table: multiply each volume driver by its time estimate, sum across all workflows, and add that total to the visible overflow from Step 1. The combined figure is the raw estimate of the full administrative demand the practice actually has. In most practices with meaningful AR backlog or recall neglect, that number is substantially higher than what shows up in the payroll.

Step 3: Subtract what should be automated instead

Not all of the invisible work should be done by a person. The automation priority framework (see the dental practice automation priority resource) identifies which workflows are better handled by automation: recall outreach through a sequenced campaign, benefit-expiry messages, and confirmation ladders are common examples of work that appears in the invisible work total but belongs to automation.

Subtract the automatable work from the Step 2 total. The remaining figure is the work that genuinely requires a person: judgment calls, exceptions, relationship-dependent communication, and complexity that automation cannot handle reliably. If a large portion of the invisible work is automatable, the first investment is not a VA but activation of tools the practice already owns.

Step 4: Add management overhead

In month one, expect to invest 2 to 4 hours per week in the engagement: daily check-ins, output review, feedback, exception handling, and process correction. This is not optional overhead. It is the investment that makes the engagement work. Practices that skip it in month one typically report poor results at month two, with no way to diagnose whether the problem was the VA, the process, or the absence of feedback.

After the first 90 days, this typically drops to approximately 1 hour per week once the VA is running established processes with minimal exception handling. Add the appropriate management overhead to the estimate based on where in the engagement timeline the practice expects to be.

The full need calculation: visible administrative overflow, plus invisible work, minus automatable work, plus management overhead.

Match the hours to the working window

Total hours are only half the specification. State when those hours must happen in the practice’s local time.

• Patient coordination needs live coverage during call peaks, lunch, and practice hours.
• Recall and treatment follow-up need patient-answering windows, often later in the day.
• Insurance and AR roles need enough overlap with payer phone lines.
• Back-office reporting and data work can happen outside practice hours and be ready before opening.

Do not write “some overlap.” Write the actual days and hours.

After calculating task time, add a 20% to 30% planning allowance for payer hold times, system delays, internal communication, breaks, and normal exceptions. A role sized to exactly 100% of measured task minutes has no room for real work conditions.

The sizing bands

Estimated need	Recommendation
Under 10 hours per week	Automation only, or a shared/part-time arrangement if human judgment is required
10 to 20 hours per week	One part-time scoped role
20 to 35 hours per week	One full-time role with a clearly defined scope
35 to 60 hours per week	Two scoped roles, not one person doing two jobs
Over 60 hours per week	A small virtual team with an internal or virtual operations lead

Under 10 hours per week: the volume does not justify a dedicated individual. The automation-only path resolves most of the need. If human judgment is required for the remaining work, a shared or part-time arrangement is appropriate rather than a full engagement.

10 to 20 hours per week: one part-time scoped role, owning one function well. One function owned well is more valuable than half of two functions owned poorly. Resist the impulse to fill the hours with variety across multiple functions.

20 to 35 hours per week: one full-time role with a clearly defined scope. This is the range where most general practices with any meaningful AR discipline issues and recall neglect will land when the invisible work is included in the estimate. The number is almost always higher than the owner’s intuition before running the calculation.

35 to 60 hours per week: two scoped roles, not one person doing two jobs. The combination of two high-volume functions into a single role is one of the most reliable paths to mediocre results in both. The person is too split to build depth in either function. When volume reaches this range, the second hire should be a new scoped role, not an expansion of the first person’s hours and responsibilities.

Over 60 hours per week: consider a small virtual team with an internal or virtual operations lead. At this volume, an unmanaged VA is insufficient. The management overhead itself requires a dedicated resource. An operations lead who manages other VAs and owns the process layer is the appropriate structure.

Start small and expand on results

Starting small and expanding on demonstrated results outperforms starting large.

Starting large means committing to a cost and a scope before the practice knows whether the model works in its specific environment. Starting at the low end and expanding on proof produces a cleaner measurement: one function’s improvement is attributable to the VA’s work, not mixed with three other simultaneous changes. If something is not working, the practice can identify and correct it at small scale rather than untangling it from a large engagement.

A practice that starts with 15 hours of scoped insurance VA work, proves the result, and expands to 30 hours six months later has de-risked both the initial and the expansion commitment. The practice knows what it is buying at each stage, because it has already seen it work.

A practice that starts at 40 hours hoping for comprehensive administrative support has no clean measurement for months and no basis for a course correction if something is not working. The signal is buried in the noise.

Start at the low end of the appropriate sizing band. Expand when the result is measurable and clear, not when the calendar hits a predetermined date.