Virtual Assistant KPI Scorecard: Metrics and Review Template

What makes a virtual assistant KPI useful?

A key performance indicator should describe an outcome the assistant can influence and the manager can verify. Examples include the share of assigned tasks finished by the agreed deadline, the number of records corrected after quality review, or whether customer requests were routed within the promised response window.

A high count of messages sent is not automatically a positive result. If every reply triggers another question or the CRM becomes less accurate, activity has increased without adding business value. Measuring work well means defining what 'complete' means and checking the quality of the finished result.

A useful scorecard has only a few metrics, a reliable source for each number, a person who reviews them, and space to explain exceptions. The manager still owns the clarity of the work, approvals, staffing capacity, and access to the tools needed to deliver the outcome.

Which KPIs should every virtual assistant manager consider?

The table below provides possible measures, not required targets for every role. Choose the measures that match the work and define the reporting period. If the denominator is extremely small, show the counts beside the percentage so one unusual case does not distort the conversation.

Always document exclusions. A task waiting for client approval should not automatically count as an assistant delay if the assistant escalated it correctly. Likewise, work marked complete but returned for corrections should be visible in the quality measure.

How do you calculate virtual assistant performance metrics?

Imagine an administrative assistant had 40 tasks due this week and finished 36 by their agreed deadlines. On-time completion is 36 divided by 40, or 90%. Suppose a reviewer checked 20 of the completed tasks and requested corrections on two. The observed rework rate in that reviewed sample is 2 divided by 20, or 10%. These numbers are hypothetical examples, not published performance benchmarks.

Now add context: three late tasks were waiting for manager approval and were escalated on time, while one had no advance warning. Counting all four as the same failure would hide the distinction between a process bottleneck and an execution gap. Keep the original metric, but record the reason categories before taking action.

Percentages should never conceal the volume of work. A 100% result based on one task is not as informative as a consistent result over dozens of comparable tasks. Where practical, show a rolling trend and an example of the actual deliverable reviewed.

• Tasks due: 40; completed on time: 36; on-time completion: 90%.
• Tasks reviewed: 20; returned for correction: 2; sampled rework rate: 10%.
• Late-task categories: waiting for approval, unexpected outage, unclear scope, or missed ownership.
• Coaching action: clarify the approval workflow and require an earlier status update for at-risk work.

How should KPIs change by virtual assistant role?

Different roles create value in different ways. A customer-support assistant may be evaluated on resolved inquiries and appropriate escalation; an executive assistant on calendar accuracy, briefing readiness, and timely follow-up; an ecommerce assistant on order exceptions and listing quality. A single company-wide 'tasks per hour' quota is unlikely to capture these differences.

Use a mix of quantity, quality, timeliness, and judgment. An assistant who handles fewer sensitive cases correctly may contribute more than someone who closes many tickets without proper checks. Assess whether the employee or contractor controls the process being measured.

What should be included in a VA performance spreadsheet?

The accompanying CSV template gives each metric a row with the definition, review period, agreed target, observed result, evidence source, exception notes, review owner, and next action. It is deliberately editable rather than pretending every business has the same benchmark. Open it in a spreadsheet tool and replace the example categories with those used in your workflow.

Record the evidence source before you begin. For task completion, that could be a project board with due dates. For response windows, it could be your ticketing system. For record accuracy, use a defined sample checked against original documents. If a metric relies only on memory or an undocumented estimate, mark it as provisional.

The best version of this template takes a few minutes to review. A sprawling dashboard with dozens of columns can become another job for the person you hired to reduce administrative work.

• Metric and plain-English definition, including exclusions.
• Weekly or monthly reporting period and source of the raw counts.
• Agreed target, actual numerator and denominator, and observed result.
• Exception notes and one named person responsible for the next action.
• Review date so improvement can be checked rather than assumed.

How often should you review a new virtual assistant?

During the first month, use frequent conversation and work examples rather than treating a dashboard as an automatic pass-or-fail system. Newly assigned work can look slower because the assistant is documenting instructions, clarifying priorities, and learning company-specific approvals.

For an established arrangement, a short weekly operational review and a broader monthly process review may be sufficient. Increase the cadence for unusual changes, not as a default form of surveillance.

1. Before the first task, select two or three measures and show a completed example of acceptable work.
2. In the first week, review samples daily where the risk or learning curve warrants it, and correct unclear instructions immediately.
3. In week two, compare the quality of output with the first-week baseline and identify recurring exceptions.
4. In week three, give the assistant ownership of a slightly larger workflow only when the existing process is stable.
5. At day 30, review the scorecard, ask what blocks progress, and revise scope or targets based on evidence.

What virtual assistant metrics can become misleading?

Metrics can backfire when they reward the wrong behavior. Counting emails sent can encourage unnecessary follow-ups. Rewarding ticket closure speed without checking resolution can push unresolved problems back to customers. Measuring attendance or keyboard activity does not prove that the important work was accurate or useful.

Consider whether the assistant can genuinely control the result. A lead-generation assistant may maintain clean contact records and timely qualified handoffs, but they cannot guarantee sales that depend on pricing, product fit, and a client's closing process. Separate controllable actions from downstream company outcomes.

Explain the purpose and scope of monitoring, minimize unnecessary personal information, and review applicable workplace privacy requirements. A scorecard should help people work more clearly, not replace management judgment with a decontextualized ranking.

What should you do when a KPI declines?

Look at the underlying examples before changing pay, staffing, or responsibilities. A decline can result from a broken tool, changing task complexity, conflicting instructions, insufficient capacity, or mistakes by the assistant. Ask which inputs changed and whether the expected standard was actually communicated.

Choose one corrective action: clarify the SOP, simplify an approval chain, provide focused training, or redefine the work queue. Review the same measure in the next period. Recording a specific change creates learning; blaming a single number rarely does.

If you are preparing a new hire, use the scorecard definitions inside the job brief and onboarding plan. Candidates can evaluate whether the job expectations match their experience, and managers can compare evidence from a trial task to the same outcomes that matter after hiring.