Virtual Assistant Test Tasks: 5 Work Samples for Coaches

How do you find out whether a virtual assistant can handle your coaching admin before giving them access to real clients? Give a shortlisted candidate one short, paid work sample using fictional records. Test the work you actually need: a scheduling proposal, inbox triage, an onboarding check, a content draft or an invoice exception report.

The most important result is not speed. It is whether the candidate completes the task accurately, identifies missing information and respects the limits you set. Someone who produces a polished document by inventing a client detail or ignoring a no-send instruction has not passed.

Below are five virtual assistant test tasks for coaches, with sample inputs, candidate instructions and an answer key for your review. Choose the relevant sample rather than assigning all five. This is a work-sample pack, not a general hiring guide or a substitute for references, interviews and lawful engagement terms.

Cover: AI-generated editorial illustration of a coach preparing an administrative handoff.

Set up a fair paid test before choosing the task

Agree the fee, expected effort, deadline and payment date in advance. Pay for the agreed assessment whether or not you hire the person. Keep the exercise short and relevant, and offer an appropriate route for accessibility needs or timing adjustments.

Use the same core brief and review criteria for candidates you are comparing. Do not turn the assessment into an endurance competition or ask applicants to produce a week of useful unpaid business work. A concise sample should answer a specific hiring question.

Choose a task that is repeatable, reviewable and safe to get wrong. All names, messages and amounts in the examples below are invented. Put “TEST DATA” on the sample documents, use no production credentials and make external actions impossible. A document or sandbox is enough.

State the tools policy too. If you permit AI assistance, spell out what may be entered and ask candidates to disclose how they checked the result. If you need to assess unaided writing, explain that beforehand. Do not hide the rule and penalize people afterward.

1. Scheduling test: resolve a booking request without guessing

Use this when: the role includes calendar coordination or session logistics. The exercise should test constraint handling, not whether the candidate knows your favorite calendar interface.

Fictional input pack

• The coach has two available appointments: Tuesday at 10:00 UTC and Thursday at 15:00 UTC. Each is 45 minutes.
• Jordan asks for Tuesday at 10:00 UTC.
• Mina asks for Tuesday at 10:00 but does not state a time zone.
• The rule says appointments are proposals until the coach approves them.
• No additional availability may be opened.

Candidate brief

Prepare a proposed booking list and draft the messages needed to resolve these requests. Do not send messages or create appointments. Show any information you need before proceeding. Use the stated availability only.

What a strong submission does

It treats Jordan's request as a possible match, not a confirmed booking. It asks Mina for the missing time zone rather than assuming that “10:00” means UTC. It notices the potential conflict and leaves the coach a clear approval decision.

The candidate can suggest the other available slot where appropriate, but should not imply that it has been accepted. A good completion note separates proposed bookings, unanswered questions and actions awaiting approval.

Stop-and-review errors: confirming both people into the same slot, inventing Mina's time zone, adding an unauthorized appointment or sending a real message. Formatting differences are much less important than those boundaries.

Debrief question: “What would you do if the client said this was urgent but neither slot worked?” Look for an escalation within your policy, not a promise to open your evening.

2. Inbox test: separate administration from coaching judgment

Use this when: the assistant will sort enquiries or prepare routine replies. The exercise should show whether they recognize when a message is not ordinary administration.

Fictional input pack

• Message A: “Can you send the link to move next week's appointment?”
• Message B: “I disagree with the amount on this invoice and want a refund.”
• Message C: “I keep avoiding the exercise we discussed. What does that say about me?”
• Approved information: one test booking URL and a rule that refunds and coaching questions go to the coach.
• Authority: categorize and draft only. Nothing may be sent.

Candidate brief

Label each message, state the next action and draft a reply only where the approved information supports it. Flag decisions outside your authority. Keep the report factual and do not interpret a client's motivation.

What a strong submission does

Message A gets a logistical draft using the supplied test link. Message B becomes a dispute requiring the coach's review, without a refund promise or a demand for immediate payment. Message C is recognized as a coaching question rather than an opportunity for the assistant to offer advice.

A concise report is enough: category, proposed action and decision owner. The candidate should not create a psychological label for the fictional client or add confidential-looking material to a public tool.

Stop-and-review errors: giving personalized coaching advice, promising a refund, claiming the coach has approved something or ignoring the no-send boundary. If the candidate drafts a short holding message for B or C, evaluate it against the brief; asking whether a holding reply is allowed can be a good clarification.

Debrief question: “Which message required a decision rather than a template?” This distinguishes recognition of an exception from simply following the order of the list.

3. Onboarding test: report readiness without reading private notes

Use this when: you need someone to check whether new clients have completed administrative steps. The task tests reconciliation and restraint, not suitability assessment.

Fictional input pack

• Alex: agreement received; access email delivered; first appointment not booked.
• Sam: agreement missing; access email delivered; first appointment booked.
• Rae: agreement received; access email bounced; first appointment booked.
• The checklist requires all three administrative steps before the readiness report is marked complete.
• There are no intake answers or coaching notes in the pack.

Candidate brief

Create a readiness report with the missing step and next action for each person. Draft one reminder using the approved wording. Do not mark a client complete because an appointment exists, and do not invent reasons for an incomplete step.

What a strong submission does

It identifies Alex's booking gap, Sam's missing agreement and Rae's failed email delivery. It proposes the correct administrative next action without deciding that anyone is unmotivated, unsuitable or unwilling to participate.

For Rae, the candidate should flag delivery troubleshooting or confirmation of the address through an approved route. Repeatedly sending the same message to a known failed destination is not a completed handoff.

The Team screenshot illustrates a real Coachful demo workspace, not a requirement to grant a candidate access. Run this exercise with a simple fictional list. Before a later live engagement, verify the permissions actually available; do not assume a screenshot proves a particular restricted assistant role exists.

Stop-and-review errors: marking everyone complete, overlooking the bounced message or requesting full session notes to perform a status check.

Debrief question: “What evidence would you need to change Rae's status?” Look for a verifiable completion event rather than confidence that the problem probably resolved itself.

4. Content test: prepare a draft without inventing expertise

Use this when: the role includes formatting newsletters, preparing captions or repurposing teaching material. Test fidelity to an approved source, not whether the candidate can imitate a generic marketing voice.

Fictional input pack

Provide this short teaching note: “Before a difficult feedback conversation, write down the behavior you observed, its effect on the work and the question you want to ask. Do not assume you know the other person's intention.” Add an approved audience description and a test resource URL.

State that the exercise contains no client case study, outcome statistic or testimonial. The requested asset is a short email draft, not a live campaign.

Candidate brief

Turn the teaching note into a concise email draft for newly promoted managers. Preserve the advice, include the supplied resource link and mark anything that needs clarification. Do not invent a client story, make an outcome claim or schedule the draft.

What a strong submission does

It keeps the distinction between observed behavior and assumed intention. It addresses the intended reader and uses the approved link. It does not transform a practical preparation exercise into a promise that difficult conversations will always go well.

Judge the draft against a small checklist: faithful substance, appropriate tone, correct link and no unsupported claim. Perfectly matching your personal punctuation preferences is not the same as demonstrating sound editorial judgment.

The Coachful Email Broadcasts screenshot illustrates where draft campaigns can be prepared. Preparation, approval and sending remain separate permissions. The test needs no email account, audience list or social login.

Stop-and-review errors: fabricated testimonials, invented results, publishing instead of drafting or uploading private material to create a more persuasive example.

Debrief question: “What did you deliberately leave out because the source did not support it?” That can reveal more than asking whether the candidate considers themselves a good writer.

5. Invoice test: identify an exception without moving money

Use this when: the role includes invoice tracking or reminder preparation. Do not use a payment account for the assessment.

Fictional input pack

• Invoice TEST-101: unpaid, due yesterday, no dispute recorded.
• Invoice TEST-102: unpaid, due yesterday, disputed by the purchaser.
• Invoice TEST-103: marked paid in the invoice list, but an attached test note calls it unpaid.
• Authority: prepare an exception report and reminder drafts only.
• Rule: no refunds, new charges, invoice edits or changes to payment terms.

Candidate brief

Identify which record can receive the standard reminder draft, which needs a dispute decision and which needs reconciliation. Explain the evidence you would seek. Do not change a record to make conflicting information disappear.

What a strong submission does

TEST-101 gets a standard reminder draft. TEST-102 is withheld from that routine workflow and escalated as a dispute. TEST-103 is flagged for reconciliation; the candidate does not assume that the note or the status is automatically correct.

A good request for evidence might ask for the authoritative payment record or a clarified reconciliation policy. The candidate should be comfortable reporting an unresolved discrepancy instead of presenting a falsely tidy result.

Stop-and-review errors: promising a refund, recommending a charge to a stored card without authorization, sending a disputed reminder or editing the conflicting status without evidence.

Debrief question: “What can you complete now, and what must wait?” The answer should distinguish useful progress from overstepping.

Use one review rubric rather than a general impression

Before sharing the task, define what you will review. You can mark each dimension as “met,” “needs clarification” or “not met,” with a specific observation. This is an editorial assessment method, not a validated psychometric instrument or a prediction of long-term performance.

• Accuracy: did the output match the supplied records and constraints?
• Clarification: did the candidate identify missing or conflicting information?
• Authority: did they respect draft-only, no-send and no-money rules?
• Completion reporting: did they distinguish finished work from blockers and approvals?
• Usability: can you understand and review the result without reconstructing it?

Record evidence rather than personality labels. “Asked for Mina's time zone before proposing a confirmed booking” is an observation. “Seems trustworthy” is an impression. Both a quiet candidate and a polished interviewer deserve assessment against the same job-related standard.

Do not average away a serious boundary breach because the document looks beautiful. Equally, do not treat a minor formatting preference as a disqualifying competence problem. If a brief was unclear, clarify it consistently rather than turning your ambiguity into a secret test.

Give the candidate a complete brief

Use this outline for the sample you select:

• Purpose: the specific job behavior the exercise assesses.
• Scope: the fictional input pack and requested output.
• Effort and deadline: an agreed time limit and submission time with time zone.
• Compensation: the agreed fee, payment route and payment date.
• Tools: what is permitted and any disclosure requirements.
• Boundaries: no real contacts, production access, publishing or money movement.
• Questions: where to ask and how a delayed answer affects the deadline.
• Finish: submit the artifact, unresolved questions and a short completion note.

Tell the candidate to stop at the agreed effort limit and submit what they have. A limited sample should not quietly become an unpaid evening because you added another request. Retain recruitment materials only for an appropriate period and explain any intended use of the work.

After the sample: test whether a real engagement returns time

A good sample supports a narrow conclusion: this person handled this defined task under these conditions. It does not establish long-term reliability, settle worker classification or remove the need for lawful contracts and appropriate data-access arrangements.

If you proceed, begin with a bounded paid engagement and the minimum necessary permissions. Track your briefing, review and correction time alongside the work transferred.

In the hypothetical planning example, six hours of administration move to the assistant while the coach spends two hours briefing and reviewing. Four hours of net time are released in that period. This is not a forecast, a fee benchmark or proof that every delegated task pays for itself.

For the broader first-hire decision, read our assistant and VA hiring guide. This article's distinct job is providing fictional work samples and observable answer keys before production access, not repeating the full hiring process.

Common questions about virtual assistant test tasks

Should I assign all five tests?

No. Choose the sample closest to the actual role. If a second skill genuinely needs assessment, agree a proportionate additional scope and compensation. Do not turn a shortlist into a source of low-cost production work.

Should I use real client messages if I remove their names?

Prefer fully fictional messages. Removing a name does not necessarily remove identifying or sensitive context. These samples work without exposing a real client's life, employer, payment dispute or coaching history.

What if the candidate asks questions instead of finishing?

Evaluate the questions. Asking about a missing time zone or contradictory payment status is part of the job. A useful submission can finish the supported work and clearly identify what remains blocked.

Test the judgment before handing over the login

Pick one sample, agree fair payment and review the result against the written brief. You are looking for accurate work, useful questions and respect for limits, not a candidate willing to say yes to everything.

We make Coachful. The Team, Email Broadcasts and invoicing screenshots illustrate workflows a later assistant might support, subject to verified permissions. Try the interactive Coachful demo to inspect the work you want assessed, then build a fictional test pack. No candidate needs access to your real practice to complete these exercises.

Research note: Current resources such as Kuya Solutions' VA interview guide discuss short paid work tests using sample data. The five coach-specific input packs, answer keys and review criteria above are original editorial examples. No vendor outcomes, hiring success rates or search volumes are claimed.