A take-home assignment is an unsupervised, time-boxed task a candidate completes on their own schedule and submits for review. It belongs to the wider family of work sample tests but differs in one way that matters: nobody watches the work happen, so the brief, the time cap and the scoring rubric carry the entire weight of fairness.
A work sample test is the general category: any task replicating real work, whether it runs on site, in a live session, or at home. A take-home is the unsupervised variant. A technical screen is a different animal again, usually live, usually early, and aimed at verifying a floor rather than measuring depth. The practical consequences differ. Supervised samples let you watch reasoning happen and ask why; take-homes hand you a finished artefact with the reasoning invisible unless you ask for it separately. Take-homes scale across time zones and suit candidates who freeze under observation. They also consume unpaid hours and can be outsourced or generated, so pair every take-home with a short live walkthrough where the candidate explains their choices. That conversation, not the artefact, is where most of the signal turns out to sit.
Cap it at two to three hours and design the brief so the cap is honest. State the limit in writing, say plainly that extra polish earns no extra credit, and score only what the cap allows. Candidates routinely spend four times the stated budget because they assume their competition will, so the cap only holds if the rubric backs it up. Anything substantial should be paid at a fair hourly rate for the role. Paying is not a formality: it signals that you value the candidate's time, and it kills the argument that you are collecting free work. If a task is too large to pay for, it is too large to ask for. Never use a live company problem as the assignment, whatever the intention behind it.
Write the rubric before the first brief goes out, and score blind. Strip names and identifying details from submissions, have two reviewers rate independently against the same criteria, then reconcile the two. Blind review is easier here than anywhere else in hiring, because the artefact travels without a face or a voice attached, and it removes the pull of a familiar university or employer name. Score against the criteria you published, not against what you wish the candidate had built. Keep criteria few and concrete: correctness, structure, clarity of the trade-offs stated, and handling of the ambiguous requirement you deliberately left in. Store submissions and both sets of scores on the candidate record in your [applicant tracking system](/ats) so a rejected candidate's outcome can be explained if they ask.
Predictably: people already working, people with caring responsibilities, and people holding an offer elsewhere. An unpaid multi-hour task filters on available free time before it filters on ability, and free time is not evenly distributed. Parents, carers and anyone working a second job feel it first. That does not make take-homes wrong, but it makes the design choices load-bearing. Offer a live alternative for candidates who prefer one. Give a generous window, at least a week, so the task can fit around a life. Do not send an assignment to somebody you have barely engaged with, since abandonment peaks when the candidate has no reason yet to invest. Watching funnel drop-off at this stage in your [recruiting analytics](/recruitment-analytics-software) tells you quickly whether the brief is too big.
Pitch N Hire unifies sourcing, screening and hiring decisions on one AI-native platform. Book a quick demo on your real roles.
Prefer to talk? Book a demo · View pricing
Free 1-user plan · No credit card · Talk to a real hiring expert
See your true cost-per-hire and how much Pitch N Hire could save you — our free Recruitment ROI Calculator gives you the numbers in under a minute. No signup required.
Open the free ROI calculatorPrefer a tailored walkthrough on your real roles? Drop your work email:
★ Free 1-user plan · No spam · Talk to a real hiring expert
