How to Manage a Dental VA's Performance

Manage output and exceptions, not activity. Screenshot monitoring and keystroke tracking produce compliance theater, damage the relationship, and correlate poorly with results. If the metrics are right, activity monitoring is unnecessary; if the metrics are wrong, activity monitoring will not save the engagement.

The reason activity monitoring feels useful is that it creates the impression of control. You can see something happening. But what you are seeing is that something happened, not that something useful happened. A VA who sends 200 emails per day and a VA who sends 40 but converts 30% of them into booked appointments are not performing the same function. Activity monitoring measures the 200. It has nothing to say about the 30%. The only measure that matters is the result, and results come from outcome metrics, not activity logs.

The Three-Layer Measurement Model

Performance measurement in a dental VA engagement has three layers, each answering a different question. All three are necessary. Any single layer on its own is insufficient.

Layer 1: Output Metrics (Did the Work Happen?)

Output metrics count activity: verifications completed, claims submitted, calls made, messages handled, reports delivered on time. These are necessary to confirm that the workflow is running and the volume is appropriate to the scope.

They are not sufficient on their own. High output with poor outcomes is a scoping problem, not a performance problem. A VA completing 80 insurance verifications per week who never escalates a payer issue is producing volume without value if the verifications are inaccurate or incomplete. Output metrics confirm that work is happening. They do not confirm that the work is working.

Layer 2: Outcome Metrics (Did It Produce the Result?)

Outcome metrics measure whether the engagement is producing its intended result: AR over 90 days, clean claim rate, open chair time, recall reappointment rate, speed to first response, unscheduled treatment recovered.

These are the metrics the engagement should actually be judged on. If the engagement was hired to improve recall reappointment rates and recall reappointment rates are not improving, the engagement is not working, regardless of how many calls are being made. Conversely, if the engagement was hired to reduce claims denials and clean claim rate is trending up, the engagement is working, even if the output volume looks modest.

Outcome metrics require a baseline to be meaningful, which is why recording baseline metrics before Day 1 is a Week -1 requirement. Without a baseline, you cannot tell whether the outcome metric is improving.

Layer 3: Quality Metrics (Was It Done Well?)

Quality metrics answer what outcome metrics miss: not just whether the result happened, but whether it happened in a way that is sustainable and accurate.

An audit sample accuracy check reviews a sample of completed work against the SOP’s definition of correct. Ten verifications pulled at random and checked against the standard will tell you more than a count of 200 verifications completed.

Escalation appropriateness tracks whether the VA is escalating the right things: both over-escalation (bringing every minor question to the supervisor) and under-escalation (handling compliance-sensitive situations independently) are signals that calibration is off.

Documentation completeness confirms that completed work is documented in a way that creates a record. A claim submitted without documentation of the verification creates a gap that surfaces in an audit or a denial dispute.

Patient complaints and compliments, where the VA is in a patient-facing communication role, track quality in the most direct possible way.

The Review Cadence

Weekly: output plus exceptions plus blockers. The purpose of the weekly review is to catch problems early, not to create a meeting for its own sake. If the VA is working correctly, the weekly meeting surfaces one or two exceptions and nothing else. If the VA is blocked somewhere, the weekly meeting is where the block is identified before it costs a week of misdirected work.

Monthly: outcomes versus targets. The monthly review is where the engagement is evaluated against its purpose. Are the outcome metrics moving in the right direction? If not, what is the diagnostic? Is the workflow correctly designed? Is the scope matched to the problem? Is there a training or documentation gap that is limiting results? Monthly is early enough to course-correct before the problem compounds, and late enough that outcome metrics have had time to show meaningful movement.

Quarterly: quality audit, scope review, role evolution. The quarterly review covers three questions. First, quality: pull the audit sample, review escalation patterns, check documentation completeness. Second, scope: has scope crept from the original written definition? This is the moment to ask directly and reestablish the boundary if it has. Third, role evolution: is the current scope still the right scope? Has the practice changed, or has the VA demonstrated capacity for expanded responsibility?

Annually: compensation, expansion, structural review. The annual review is the moment to make structural decisions: whether the engagement continues, whether the scope should formally expand, whether the compensation structure should change.

Track Whether Management Time Is Falling

The relationship has not produced its full return while the manager still checks every item. Track the manager’s supervision hours each week from the start. Include time spent assigning, answering routine questions, reviewing, correcting, and chasing updates.

That number should fall as each workflow becomes reliable. If output rises but supervision time does not fall, look for the reason. The standard may be unclear, the procedure may be incomplete, decision rights may still be too narrow, or the manager may not have defined when review can decrease.

Do not judge trust as one general feeling. A VA may need no review on recall calls and full review on a new denial workflow. Record the review rate separately for each type of work.

Reduce Review by Evidence

Move from full review to random sampling in planned steps. A practical sequence is 100 percent, 50 percent, 20 percent, then exception reporting with occasional spot checks. Set the accuracy requirement and sample size before moving to the next step. Higher-risk work should have a stricter standard.

If an error appears, tighten review only for the affected workflow. Find out whether the cause was a missed instruction, a missing procedure branch, poor judgment, or concealment. Set a clear route back to the earlier review rate. One mistake in insurance verification should not return every other task to day-one supervision.

Diagnosing Underperformance

Diagnose before acting. The cause of underperformance is usually not effort.

When volume is low. Low volume most often means unclear priorities or access blockers, not insufficient effort. Before addressing the person, confirm that the task is clearly defined and that all system access is working as expected. A VA who is waiting for a system to load, cannot access a payer portal, or does not know which tasks to prioritize will show low volume without any failure of commitment. Fix the environment before addressing the person.

When errors are high. High error rates most often mean inadequate documentation or training, not carelessness. If the SOP does not describe what correct looks like in specific terms, the VA cannot produce it consistently. “Process the claim correctly” is not an SOP. “Enter the date of service from the clinical note, use the tooth number from the chart, verify the CDT code matches the narrative, and flag for review if the code is not on the payer’s fee schedule” is an SOP. If the SOP is that specific and errors are still high, then the training is the next place to look. Fix the SOP before addressing the person.

When questions are constant. Constant questions are not an attention problem. They are a documentation gap. Every question the VA asks that is not already answered in the SOP is a missing section. Treating constant questions as interruptions means the same questions recur indefinitely, because the gap that generates them is never closed. The exception log is the mechanism that converts those questions into SOP improvements. Log every question, add the answer to the SOP, and the question stops recurring. A practice that treats questions as problems to be managed rather than documentation signals to be addressed will always have a VA that asks too many questions.

When metrics are flat despite high activity. Flat metrics with high activity mean either the wrong workflow was assigned (the VA is working on something adjacent to the problem rather than the problem itself), or the underlying constraint lies somewhere else in the practice that the VA’s work cannot reach. A VA making recall calls into a pool of patients who are unscheduled because the practice is three months out on new appointments will not move the recall reappointment rate. The constraint is the schedule, not the calls. Re-diagnose the underlying problem before changing what the VA is doing.

When scope is confused. Scope confusion means scope has crept. Return to the written scope document from the original agreement and re-establish the boundary explicitly. The conversation is not about what the VA is capable of. It is about what the engagement was designed to produce and what adding unplanned tasks does to the original problem. Adding capacity requires expanding the engagement formally. It is not the same as adding tasks informally.

When there is team friction. Team friction means the internal integration described in the onboarding process was never completed or was not effective. This requires an owner-level conversation with the team, not just the office manager managing it as an interpersonal issue between team members and the VA. The owner needs to re-establish why the VA is part of the team and what success looks like for everyone.

The Practice Side of the Equation

In a substantial share of “the VA isn’t working out” cases, the diagnosis points back to the practice.

The practice did not document the process before hiring, so the VA has been operating from incomplete information. Scope was never defined, so the VA has been doing whatever came up rather than owning a specific workflow. The wrong function was chosen for the problem: an insurance specialist was hired to solve a scheduling problem, or a scheduling coordinator was hired when the real issue was accounts receivable. The underlying problem was demand or clinical capacity, which no VA can address because it lives upstream of administration.

When the diagnosis points to the practice, the right response is to fix the practice side, not to replace the VA. Replacement without diagnosis reproduces the failure. The next VA will arrive to the same undocumented process, the same undefined scope, the same mismatched function. They will fail for the same reason the previous one did, and the conclusion will be that “VAs don’t work for us” when the accurate conclusion is that the structure of the engagement has never been designed to work.

A platform or advisor willing to name this and help fix the practice side is worth more than one that substitutes a new candidate. The candidate is not the variable. The structure is.

When to End an Engagement

Some engagements should end. The question is whether the reason is genuine or whether it reflects a preparation failure that would be better addressed than abandoned.

Genuine cause to end an engagement: Repeated compliance or accuracy failures after documented remediation has been provided. Dishonesty about qualifications, experience, or completed work. Inability to reach competence after adequate training and documentation have been delivered. A persistent trust breakdown where the relationship cannot be repaired.

These are grounds to end an engagement because the remediation path has been exhausted or the foundational requirement of trust is gone.

Treat conduct separately from normal underperformance. A suspected privacy or security breach, deliberate concealment, falsified work, or unexplained absence may require immediate access removal and a formal investigation. Preserve relevant records and obtain professional advice where reporting, employment, privacy, or contractual duties may apply.

Not grounds to end an engagement: The practice never documented the process, so the VA was operating without clear instructions. Scope was never defined, so performance cannot be fairly evaluated. The wrong function was chosen for the actual problem the practice needed to solve. The underlying problem was demand or clinical capacity, which is a business constraint the VA cannot address.

For a performance problem, write a short recovery plan before deciding to replace. Name the affected workflow, observed evidence, likely cause, action on both sides, review rate, success measure, and checkpoint date. If the person succeeds, restore the prior operating level. If the same gap remains after adequate support, the practice has a documented basis for its next decision.

See Also

• Why Dental VA Engagements Fail, the ten failure modes and the diagnostic that identifies which one is at work
• The Dental VA Onboarding System, the onboarding system that sets the performance baseline and establishes the review cadence
• The Dental VA Training System, the three-layer training model and how the exception log converts ramp into documentation

Ending an engagement for these reasons attributes to the VA a failure that belongs to the preparation. It also produces the same outcome as the diagnosis-free replacement described above: a new person in the same broken structure, failing for the same reason. The right response to a preparation failure is to fix the preparation, either with the current VA if the relationship is intact, or with a clearly defined structure before hiring again.