How to Vet a Virtual Assistant: Test the Work Before You Commit

Published September 18, 2026.

Resumes and interviews reward people who present well, not people who work well. Vetting flips that order. You give a candidate a small, realistic piece of your actual work and watch how they handle it, then you decide on evidence rather than impression.

how to vet a virtual assistant

Vetting a virtual assistant means running a short, paid trial of the exact work you plan to delegate, scored against criteria you wrote down before the candidate started. The goal is to predict performance on your recurring tasks, not to test general intelligence or loyalty. A good vetting process has four parts: a defined sample task, a rubric, a communication check, and a reference check.

Keep the whole exercise to a few hours of candidate time and a small fixed payment. You are buying signal, not free labor. When the sample mirrors the real queue, the result tells you whether the person can take work off your plate without creating a new review burden.

A useful rule is to vet the task family, not the person in the abstract. If you need inbox triage and meeting follow-up, the sample should be inbox triage and meeting follow-up. If you need research summaries, the sample should be a research summary. Generic logic puzzles and personality quizzes do not transfer to the daily work.

Separate screening from vetting

Screening answers a narrow question: does this person clear the basics? That means availability, time zone overlap, rate range, tools they already know, and written English at the level your work requires. You can screen in twenty minutes with a short form and a couple of written questions. Most candidates should be removed here, before you invest in a work sample.

Vetting answers a different question: can this person do the work to your standard, and can they communicate about it? That requires real output. Do not merge the two steps. When screening and vetting collapse into one long interview, you end up judging conversational skill and calling it capability.

The sequence matters. Screen wide and cheap, then vet narrow and carefully. If you cut the funnel in the wrong order, you spend your most expensive attention on people who were never available or never suited to the role.

Build the work sample from your real queue

Choose one task you delegate or intend to delegate within the next month. Sanitize anything sensitive, then write a short brief with the context a new hire would actually receive. Attach the source material and state what a finished deliverable looks like. Set a time box, usually ninety minutes to two hours, and pay the candidate for it.

A strong sample has three qualities. It is representative, meaning it looks like the work that will recur. It is bounded, meaning the candidate can finish it inside the time box. And it is judgment-bearing, meaning at least one decision requires the candidate to choose rather than copy.

For an inbox triage sample, provide ten redacted emails and ask for a sorted queue with a one-line reason per message and a list of items needing your decision. For a research sample, supply a question and ask for a one-page brief with sources and a stated confidence level. For a calendar sample, provide scheduling constraints and ask for a proposed week with conflicts flagged.

Score the sample with a written rubric

Write the rubric before you read any submission. Once you have seen a candidate's work, it is very hard to judge it against criteria you invent afterward. Fix the dimensions and the scoring scale in advance, then score each candidate on the same sheet.

Score each dimension on a simple one-to-five scale and write one sentence of evidence per score. The written evidence matters more than the number. It forces you to separate what you observed from how you felt about the candidate, and it gives you something concrete to discuss in the next conversation.

Test communication and judgment, not just output

Two candidates can submit identical deliverables and still be very different hires. The difference shows up in how they handle the parts of the brief that were unclear. Look for a candidate who names the ambiguity, states the assumption they made, and says what they would do differently with more context.

Send one asynchronous follow-up after the sample. Ask a question that requires them to explain a choice, not restate the task. A strong answer is short, direct, and honest about limits. A weak answer is long, defensive, or padded with generic process language. You are checking whether you can correct and direct this person over text, because that is how much of the working relationship will run.

Also watch how they treat the time box. Someone who blows past it without a note either cannot estimate their own pace or will not tell you when a task is expanding. Someone who finishes early and says so is giving you useful information about capacity.

Check references and verify claims

Ask for two references who managed the candidate's work, not friends. Ask each reference one focused question: describe a task this person owned end to end and how it went. Listen for specifics. A reference who can only offer praise without an example is not giving you evidence.

Verify the concrete claims that affect your decision: tools they say they know, hours they can cover, and any certification that matters for the role. You do not need to investigate a candidate's whole history. You need to confirm the few facts your plan depends on.

Keep records of what you asked and what you learned. If two candidates are close, the reference notes often decide it. They are also useful if the same person applies again later or if a colleague asks for your reasoning.

Keep the process fair and lawful

Vet the work, not protected characteristics. Ask every candidate in the same role for the same sample and score with the same rubric. Avoid questions about family plans, age, nationality, health, or other personal matters that are unrelated to the tasks. If you engage help through a staffing firm, the same anti-discrimination rules apply to the work you direct.

Pay for work samples and be clear that a sample is not a promise of ongoing work. This protects both sides and keeps the relationship honest from the first interaction. If a candidate asks why they are doing a paid sample, the answer is simple: you want both of you to see the real work before committing.

Frequently asked questions

How long should a vetting process take?

Plan for one short screen, one paid sample of about two hours, a follow-up message, and two reference calls. That is usually enough to make a confident decision without asking the candidate to invest an entire day.

Should I hire without a work sample if the references are excellent?

Strong references help, but they describe past work you cannot see. A short sample costs little and tests the specific tasks in your queue. Use both, and let the sample carry the greater weight for task-heavy roles.

What if the sample is good but the follow-up communication is poor?

That is a real risk signal. Clear written communication is part of the job for almost every delegated role. Address it directly and consider a second, smaller follow-up. If it still breaks down, the fit is likely wrong.

Sequence vetting after a short screen

The best vetting process is boring by design. Screen quickly, sample the real work, score against a fixed rubric, check two references, and decide. Compare this approach with virtual assistant interview questions to see which questions belong in the screen and which belong after the sample. When you are ready to run the full search, the hiring guide walks through sourcing, offers, and the first thirty days. The Small Business Administration guidance on hiring and managing employees is a useful reference for structuring a fair process.