Nothing predicts kitchen performance like a kitchen. Two hours on the line tells you what no interview can: pace, cleanliness under pressure, how someone takes a correction mid-service. That is why hospitality invented the trial shift - and why its abuse as free Saturday labour has made candidates rightly wary of the words.
- A trial shift is a work sample - the most predictive test hospitality has, when it is short and structured.
- Unpaid multi-hour trials are legally risky in most markets and reputationally toxic everywhere.
- Screen first, trial last: the trial is for two finalists, not a free-labour audition queue.
- Decide what you are watching for before they arrive, or you will just remember charisma.
When is a trial shift legal?
Rules differ by country, but the pattern is consistent: a short, genuinely evaluative work sample is broadly defensible, while productive work that benefits the business typically must be paid - and in several markets any unpaid working time is simply prohibited. A full evening carrying real tables for free fails both tests almost everywhere.
- Keep it short: one to three hours demonstrates everything a trial can demonstrate.
- Pay it - at least the hourly rate, whatever the local minimum allows. It removes the legal question and signals you are serious.
- Paper it lightly: confirm in writing the time, the pay, what happens next, and insurance coverage while they are on your premises.
- Never schedule trials to plug rota holes. Candidates talk, and 'they use trials as free cover' is a review that follows you.
What should a trial actually test?
Decide before they arrive, and write it down - three to five observable things, scored while watching, not reconstructed from memory afterwards:
- Pace and economy of movement during a defined push - not raw speed, but calm efficiency.
- Hygiene and station discipline nobody has to prompt.
- How correction lands: tell them one thing mid-trial and watch the next ten minutes.
- Team radar: do they see the colleague drowning at the espresso machine?
- Guest tone, if the role faces guests at all.
Structured observation matters for the same reason structured interviews do: unstructured trials get remembered as vibes, and vibes hire confident people rather than good ones - the bias mechanics from blind screening apply on the kitchen floor too.
Who earns a trial invitation?
Finalists, not applicants. The trial is expensive for both sides - your team's attention, the candidate's free morning - so it belongs at the end of the funnel, after cheap screening has done its work. A two-minute scored quiz filters availability, experience signals and judgement first, as laid out in the quiz screening guide; then the top two cook.
This sequencing also fixes the ghost-trial problem: candidates skip trials they suspect are auditions for nothing. An invitation that arrives after visible, respectful screening - with pay stated - gets accepted.
The takeaway
Short, paid, structured, finalists-only, with next steps and a date attached. Run that way, the trial shift is the best predictor in hospitality. Run any other way, it is a lawsuit with a side of bad reviews.
Qwiza handles everything before the apron: scored screening, ranked shortlists and interview booking - so the two people you trial are the two worth your kitchen's time.
Trialing ten to find one?
Qwiza's scored quiz filters the field first, so your trial shifts go to two real finalists instead of a casting call - 48-hour pilot target.
See how Qwiza worksFrequently asked questions
Are unpaid trial shifts legal?
In most European markets, genuinely productive work must be paid, and several jurisdictions bar unpaid working time outright; only a short, clearly evaluative sample sits in defensible territory. The safe rule: keep trials under a few hours and pay them - it ends the legal question and doubles as an employer-brand signal.
How long should a trial shift be?
One to three hours, covering one defined push - a breakfast service, a delivery window. Everything a trial can reveal shows up in that window; anything longer starts being labour, not evaluation.
What if a candidate refuses a trial shift?
Ask why before judging. Experienced candidates refuse unpaid or vague trials because they have been burned - a paid, two-hour, clearly framed trial with next steps attached gets a very different answer. Persistent refusal of even that tells you something useful too.
