Screening exists to answer one question: who should we speak to first? Done badly it becomes a filter that removes good people for irrelevant reasons. Done well it is simply a fair way of ordering a list.
A short structured quiz - five to ten questions, answered on a phone - does this better than a CV pile for frontline roles, because every candidate answers the same questions and you can compare them directly.
The four types of question worth asking
- Hard requirements. Licences, certificates, legal right to work, availability for the actual shift. These are pass or fail, and they belong first so nobody wastes five minutes on a job they cannot take.
- Experience, expressed concretely. Not "do you have experience" but "which sections have you run alone" or "how many drops a day have you handled". Specific answers are hard to bluff.
- Situational judgement. One or two short scenarios from the actual job with plausible options. "A guest complains their order is wrong during the busiest hour. What do you do first?" You are looking for priorities, not correctness.
- Motivation and fit, briefly. One open question, read by a human, never scored automatically. "Why this job, right now?" answers are the most useful thirty seconds in the whole process.
How to weight answers
- Grade options, do not just count them. Each answer option is worth something: excellent, good, acceptable, weak. That turns a quiz into a ranking rather than a pass mark.
- Keep hard requirements separate from the score. Missing a legally required licence is not "minus ten points", it is a different category. Mark those questions as must-pass and flag the answer instead of burying it in a total.
- Use completion time as a signal, not a verdict. Someone who answers ten questions in forty seconds probably did not read them. Someone who took twelve minutes may have been on a bus. It informs the ranking; it never decides it.
- Never score the open answer automatically. It exists for a person to read.
Where fairness is won or lost
Three rules keep a screening quiz defensible:
- Ask only what the job needs. Every question should map to something the person will do. If you cannot explain why a question predicts performance, delete it.
- Never reject automatically. A score orders a list; a human decides who is contacted and who is not. This is both the fair position and, in most jurisdictions, the safe one under rules on automated decision-making.
- Be able to explain a score. If a candidate asks why they were ranked where they were, you should be able to answer in one sentence. If the answer is "the algorithm", the design is wrong.
How long should it be?
Five to ten questions. Below five you learn too little to rank. Above twelve completion falls off sharply on mobile, and the people who abandon are disproportionately the ones with the least free time - often exactly the candidates you want.
What a good quiz replaces
Not the interview. It replaces the CV sift and the first phone screen: the twenty minutes per candidate spent establishing availability, licences and basic experience. The interview then starts from a real conversation instead of a checklist.
The takeaway
A screening quiz is a fairness tool as much as a speed tool: everyone answers the same questions and gets compared on the same basis. Qwiza generates one from your job description, grades each option, flags must-pass answers, and hands you a ranked list - with the final call always yours.
Hiring frontline staff?
Qwiza turns your job description into a multilingual quiz campaign on Instagram & Facebook and delivers scored, qualified candidates - 48-hour pilot target.
See how Qwiza worksFrequently asked questions
Can a quiz score reject candidates automatically?
It should not, and in many jurisdictions automated rejection triggers additional legal obligations. Use the score to order the list and to flag missing hard requirements, and keep the decision to contact or decline with a person. This is also better hiring: scores miss context that a thirty-second read of an open answer catches.
How many questions is too many?
Beyond ten to twelve, completion drops sharply on mobile, and the people who abandon tend to be those with the least time - shift workers, parents, people applying between jobs. Five to ten well-graded questions produce a more useful ranking than twenty that half your pool never finishes.


