Take-home assignments and how much is too much

Take-home tasks have spread well beyond software into marketing, analysis, design, operations, teaching, and management. Done well they are the fairest assessment available — a structured problem, evaluated on output rather than on how confidently you speak. Done badly they are a tax on candidates, and the cost falls hardest on people with caring responsibilities or a full-time job.

What a reasonable brief looks like

• A stated time budget, and one under about four hours.
• A defined deliverable, so you know when you are finished.
• Clear evaluation criteria, or at least a statement of what they care about.
• A synthetic or anonymised problem, not a live piece of the company's work.
• A conversation afterwards where you walk through your reasoning.
• Sent after at least one human conversation, not as the first step.

Missing several of these is a signal about the organisation, not just the task. A brief with no time budget and no criteria is usually a sign that the team has not agreed what good looks like, which is also how they will manage you. Take-home tasks and multi-stage processes are easier to manage when you account for task switching cost.

The unpaid-work question

Some briefs are indistinguishable from a piece of billable work: audit this real account, write this actual campaign, produce this specific deliverable for this named client. If the output could be used tomorrow without modification, you are being asked to work for free. Coursera can help identify structured courses for closing a specific skills gap.

You have three options and all are legitimate. Ask for the scope to be made synthetic. Ask to be paid for it, which some employers will do and which tells you a great deal about them either way. Or decline and offer an alternative — a portfolio piece, a shorter exercise, a live working session.

A task that produces something the company can use is not an assessment. Assessments are designed to be thrown away.

Negotiating scope

Asking for a smaller version is normal, professional, and rarely held against you. It also demonstrates something the task cannot: that you scope work before starting it.

"I can give this about three hours. Would it be more useful if I did the first two sections properly rather than all four at a sketch level? Happy to talk through how I would approach the rest."

An employer who reacts badly to that question has answered a question of your own.

Doing it well

Read the criteria and answer them

The most common failure is answering a more interesting question than the one asked. If the brief asks for a recommendation, lead with the recommendation. Reviewers usually score against a rubric, and elegance in an unrequested direction scores nothing.

State your assumptions

Every brief is underspecified, deliberately. Writing down the assumptions you made and why is often worth more than the analysis itself, because it shows how you behave when information is missing — which is the actual job.

Respect the time budget, visibly

Do not spend twelve hours on a four-hour task. Reviewers can usually tell, and it is scored badly: it signals poor judgement about effort, and it makes the comparison with other candidates meaningless. Better to stop at the budget and add a short note: "With more time I would have tested X and validated Y." That note is frequently the strongest part of a submission.

Make it easy to review

A short summary at the top — what you were asked, what you concluded, what you would do next. Reviewers are reading several of these in an evening, and clarity at the top buys you a more generous read of everything below.

The follow-up conversation

If there is a walkthrough, that is where most of the assessment actually happens. Be ready to explain why you chose the approach, what you would do differently, where you are uncertain, and what you would need to do it properly. Defensiveness about the weak parts is the most common way a good submission turns into a rejection; naming the weak parts first is the most reliable way to be trusted.

When to decline

Decline when the brief asks for more than a day, when it arrives before any human contact, when it is clearly live work with no offer of payment, when the criteria cannot be explained on request, or when you simply do not have the hours and a smaller version has been refused. Say it plainly and without apology — that you are glad to demonstrate the work in another format. Some processes end there. That is information about how they treat people, delivered early and cheaply.