Test tasks: what we look for and how to approach them

Test tasks divide opinion, and reasonably so. A badly designed one wastes a candidate's evening and tells the studio little; a good one shows more about how someone works than any conversation could. We use them because, done honestly, they reveal the thing interviews cannot: how a person reasons when the problem is real and the time is short. This is how to read one, and what a strong answer usually shows.

What a test task is really for

A test task is not an exam with a right answer. It is a small, honest sample of the work itself, designed to show how a candidate thinks through ambiguity, what they prioritise and how they communicate their choices. The finished artefact matters less than the reasoning that produced it.

That framing changes how a task should be approached. Someone trying to guess the answer we want will produce something safe and forgettable, while someone who reasons openly and explains their trade-offs will show us far more, even if their answer differs from the one we imagined.

Reading the brief

Most test tasks are deliberately a little ambiguous, because real briefs are. The first useful thing a candidate does is decide what the task is asking and state that reading up front, so the reviewer can follow the choices that come after. Naming your interpretation is not a weakness; it is the clearest sign that you understand the work.

• Write down what you think the task is asking.
• Note the assumptions you had to make.
• Decide what you will deliberately not do.
• Say how you would spend more time if you had it.

These notes are often the most valuable part of a submission. They show a candidate who can turn a vague request into a defined piece of work, which is a large part of every job in a small studio.

Prioritising under a limit

Every good test task has more in it than the time allows, and that is on purpose. The interesting information is in what a candidate leaves out. Someone who tries to do everything will do it all thinly; someone who chooses a few things and does them well shows judgement, which is the quality we are actually assessing.

We would rather see one part done thoughtfully with a clear note about the rest than a shallow pass at the whole. A candidate who says here is what I did, here is what I would do next, and here is why I chose this order is showing us how they would behave on a real project with a real deadline.

What we look at

The table below is how we weigh a submission, from the most important signal to the least. It reflects the habits that matter in a small team, where clear reasoning is worth more than polished output.

Writing about your choices

How a candidate writes about their own work is a large part of what we learn. A short, honest explanation of a decision, including why the alternatives were rejected, tells us more than a polished deliverable with no commentary. We are looking for someone who can think in public without defensiveness.

That means naming the weak parts of your own work before we do. A candidate who says this section is the weakest and here is why is showing exactly the self-awareness that makes someone easy to work with and difficult to fool. Confidence is common; honest self-assessment is not.

Making the task fair

We try to design tasks that are fair in the ways that matter: short enough to fit in an evening, clear about what is wanted, and free of anything that depends on private knowledge or an expensive tool. A task that only rewards someone with a lot of free time is not measuring the thing we claim to measure.

We also explain what the task is for before it starts, and we never use a candidate's unpaid work for anything but assessing their application. Fairness here is not just a courtesy; it is how we find out how someone really works, because a person who feels their time is respected will show us their genuine thinking rather than a guarded guess.

What a strong answer demonstrates

A strong test task is not the cleverest submission we receive. It is the one that shows a candidate turning an ambiguous brief into a defined piece of work, making honest choices under a limit, and explaining those choices plainly. That combination is rare, and it is a far better predictor of how someone will behave on a real project than any answer that merely matches what we had in mind.