Screen a VA candidate with a paid inbox triage sample

Set the sample before you interview

A paid inbox triage sample should be designed before the first interview so every candidate receives the same task, deadline, and evidence. Write the brief as if an existing assistant were covering the inbox for one morning, with a start state, a cutoff time, and a named reviewer. Keep the exercise short enough to finish in sixty to ninety minutes, because a longer sample mostly measures stamina. State the pay rate and the time limit in the brief so the candidate can plan the work honestly.

Choose four or five messages that require different judgments rather than five versions of the same request. Include one message that needs a calendar change, one that needs a document, one that needs a polite deferral, and one that should be escalated rather than answered. Add a message from a known contact that looks urgent but conflicts with the approved calendar. That mix reveals whether the candidate reads context or simply clears the newest item first.

Decide in advance what the candidate will receive and what the reviewer will keep. Send the fixture as a read-only folder with a short cover note that names the deadline and the person to contact if the brief is unclear. Keep the scoring sheet out of the candidate's view. A candidate who knows the exact scoring dimensions will write for the rubric instead of for the client.

Build the inbox fixture from safe records

Replace every customer name, address, account number, and price with invented but consistent data. Keep the thread structure realistic, including a prior reply, a forwarded note, and an attachment label that does not match the message. Put the fixture in a dedicated mailbox or a shared document that the candidate can open without touching production systems. Remove signatures, phone numbers, and any internal links that would expose real staff or tools.

Write a short key that tells the reviewer which messages are safe to answer, which need a draft only, and which must be routed to a named owner. The candidate should see the same key except for the routing answers. If the exercise asks for calendar work, provide a blank test calendar and a clear rule about double-booking. A fixture that depends on live access turns a skills test into a security risk.

Keep the fixture version number in the folder name so two candidates never receive different messages by accident. If you change a message between candidates, reset the comparison and note the change in the review. A small version label prevents a later argument about whether the exercise was fair. The same discipline applies to the scoring sheet and the time limit.

Score triage judgment, not speed alone

Use a scorecard with four dimensions: accuracy, priority order, communication, and escalation judgment. Accuracy covers names, dates, amounts, and whether the reply answers the actual question. Priority order asks the candidate to explain why the first item came first, not just to sort by arrival time. Communication covers tone, length, and whether the draft says what happens next.

Escalation judgment is the dimension most buyers forget to score. A candidate who spots a pricing exception, a legal threat, or a request from an unauthorized contact should stop and route it with evidence. Give credit for a well-written question when the brief is genuinely ambiguous. Penalize silent guessing more heavily than a slow but correct handoff.

Calibrate the reviewer before the first candidate by scoring one practice submission together. Compare the scores, discuss the evidence, and rewrite any rubric line that two reviewers interpret differently. A scorecard that produces a different answer for the same output is not ready to compare people. Keep the calibrated version with the exercise.

Compare candidates on the same evidence

Review the samples in the same order and with the same reviewer, then record one paragraph of evidence for each score. Do not let a strong interview answer overwrite a weak sample, and do not let a polished sample hide a missing attachment. Keep the candidate's written questions beside the output because they show how the person handles uncertainty. If two reviewers disagree, discuss the specific evidence rather than averaging the scores.

Ask each candidate to walk through one decision after the sample is submitted. Listen for whether the person can name the source, the rule, and the consequence of the choice. A candidate who says the message felt urgent has not explained the decision. A candidate who says the client's contract renewal date made the message important has given you something you can test again.

Watch for risky shortcuts

A fast sample can be a warning sign when the candidate invents a policy, promises a refund, or marks a message complete without the requested document. Check whether the candidate used the approved template or quietly wrote a new one. Look for replies that copy the client's wording without checking the underlying record. Those shortcuts often survive a trial period because nobody reads the sent folder closely.

Also check how the candidate handled the message that should not be answered. A reply that sounds helpful but commits the business to a date, discount, or scope is a serious error. The sample should make that boundary visible without turning the exercise into a trick. Tell candidates in the brief that some messages may need a draft, a question, or an escalation.

Read the sent folder for tone as well as content, because a technically correct reply can still create a commitment the client did not approve. Look for phrases that imply a guaranteed date, a discount, or a priority the brief did not grant. Ask the candidate what they would do if the client pushed back on the boundary. The answer shows whether the person understands why the limit exists.

Decide with a written recommendation

End the review with a one-page recommendation that names the strongest evidence, the main risk, and the first lane you would assign. Separate what the sample proved from what the interview suggested, because the two carry different weight. If the candidate is strong on execution but weak on escalation, say so and plan a narrower first scope. A clear recommendation makes the hiring decision easier to defend later.

Keep the sample, scorecard, and recommendation for the first review cycle after hiring. Compare the first month of live work with the sample to see which predictions held. If the sample missed a skill the role needs, change the exercise before the next hire rather than blaming the new assistant. The sample is a hiring instrument, and it improves only when you check its results.