Take-Home Tasks: What Is Reasonable and What Is Not
Some are a fair assessment and some are unpaid work or a filter on free time. How to tell, how to negotiate, and when to decline.
A take-home task can be the fairest stage in a process — it assesses what you can do rather than how you present, and it suits people who interview badly and work well. For a practical example of recording work completed over time, see online timesheets.
It can also be a filter on who has free evenings, or unpaid work with a rejection attached. Telling them apart is worth doing before you start. For an independent public reference, consult CareerOneStop interview guidance.
What a reasonable task looks like
Scoped in hours, not days. Two to four hours is common and defensible. A task described as "should take a day or two" is asking for something else.
A stated time expectation, and one they mean. "Spend no more than three hours" from an employer who then compares against submissions that took twelve is not a real limit.
Artificial or public data. A task using their actual current problem, with their actual data, is work.
Assessed by someone who explains the assessment. You should be able to ask what they are looking for.
Discussed afterwards. The best version is a task you then walk through in conversation, because that is where the reasoning is visible and where it becomes very difficult to outsource.
Appropriate to the stage. A substantial task before anyone has spoken to you is a poor use of your time and a sign of a process that does not value it.
What is not reasonable
Work they will use. Redesign this page, write this campaign, solve this live problem. If the output has commercial value to them, it is work.
Open-ended scope. No time limit and no defined deliverable, so candidates compete on hours available.
Very long tasks. Anything above about eight hours unpaid is a substantial ask, and it excludes people with caring responsibilities, second jobs or health constraints.
Short deadlines. "By Monday" issued Friday evening.
A task at first stage, before any conversation.
And several tasks across stages that were not disclosed at the start.
Asking about it
Reasonable, expected, and it tells you about them.
"Roughly how long should this take?"
"What are you assessing?"
"Is there a deadline, and is there flexibility?" A process that cannot accommodate a candidate who works full time is telling you something.
"Will we discuss it afterwards?"
"Is AI assistance permitted?" Increasingly relevant, and asking reads as taking the instructions seriously rather than as looking for a loophole. See where the line is.
If the answers are vague, that is information.
Negotiating it
More possible than people assume.
Ask for more time. "I work full time — could I return this by Thursday rather than Monday?" is granted routinely, and refusing it is a signal.
Propose a shorter version. "I can give this three hours — would you rather I did the full scope roughly, or one part properly?" This is a good question and it frequently improves what you submit.
Ask whether a live session could replace it. Some employers accept a paired or discussion-based alternative, particularly where a candidate explains why.
Ask about payment for anything substantial. Some employers pay for take-home work, more do than used to, and asking is not unreasonable for a task of eight hours or more.
And ask for an adjustment if you need one. Extra time for a disability or a health condition is an adjustment and requesting it is normal.
Doing it well
Stay within the stated time, and say what you did. "I spent three hours; with more I would have added X" is a strong submission — it demonstrates judgement about scope, which is frequently what is actually being assessed.
Do not gold-plate. A candidate who spends twelve hours on a three-hour task has answered a different question and it is visible.
Explain your reasoning. A short note covering what you assumed, what you prioritised and what you would do differently is worth more than extra output.
Address the brief. The most common failure is a technically good submission that does not do what was asked.
Ask a clarifying question if the brief is ambiguous, rather than guessing. It is assessed positively.
And submit on time, or say in advance that you cannot.
When to decline
When the scope is unreasonable and they will not move. "Thank you — I'm not able to commit that much unpaid time. I'd be glad to discuss the work in a conversation or do a shorter version." This is a legitimate position and some employers accept it.
When it is plainly their live work.
When it arrives before any conversation and you have limited time to allocate across a search.
And when the process has already asked for a great deal. You are allowed to decide the cost has exceeded the probability.
Declining costs you that process. Say so pleasantly, and understand it is a real trade rather than a costless stand.
The fairness problem
Worth naming, because it is why this stage is contested.
Take-home tasks systematically favour people with free evenings. Candidates with children, second jobs, caring responsibilities or health conditions compete against people with none of those.
Employers who do this well set a hard time limit and mean it. Employers who do it badly let candidates compete on hours and then select the ones with the most.
If you cannot spend twelve hours, say what you did in three. That is the honest submission and a well-run process reads it correctly.
The short version
Reasonable: two to four hours, scoped, artificial data, discussed afterwards.
Not reasonable: their live work, open-ended scope, short deadlines, or a task before anyone has spoken to you.
Ask what they are assessing, how long it should take, and whether AI assistance is permitted.
Stay within the stated time and say what you did with it — scope judgement is frequently what is being assessed.
And declining is legitimate, at a real cost you are entitled to weigh.