A realistic job preview shows the real shift before someone applies: hours and pay in the ad, a short video of the actual floor, situational questions that put candidates inside the job, and a candid Q&A with the people already doing it. Done properly, the wrong people leave at the quiz instead of in week one. In Starred's 2026 Hiring Benchmarks Report, built on nearly 70,000 open-text comments, expectation setting drew the highest concentration of negative sentiment of any theme: 37.8 per cent of comments mentioned it, and 74.5 per cent of those were negative.
- Expectation setting drew the highest concentration of negative sentiment in Starred's 2026 report: 37.8 per cent of nearly 70,000 candidate comments mentioned it, and 74.5 per cent of those were negative.
- A situational question written around the real hard part of the shift is a preview and a screen at once: the wrong candidates leave before you book an interview.
- Describe physical demands and hours in the ad, but never score candidates on them; grade judgement, not fitness, age or availability.
- Judge a preview by two numbers moving in opposite directions: more drop-off before the interview, fewer quits after the start date.
Why do new hires quit in week one?
Because the ad was fine and the job wasn't the ad. Glider.ai's realistic job preview guide (updated 01-09-2026), citing Corvirtus, says nearly half of new employees regret taking a job within their first week.
The ghosting data points the same way. In a 2026 Click Boarding survey of 2,000 US job changers, run through OnePoll, 53 per cent admitted ghosting a new employer and 45 per cent of those did it between accepting the offer and day one. Asked why, 39 per cent of respondents named conflicting or misleading information and 36 per cent a lack of clarity about the role. It's a vendor-commissioned survey of US workers, so treat it as a signal, not a Croatian statistic.
Starred's own summary of its 2026 data says candidate frustration has shifted away from assessments and pay towards unclear expectations. Candidates will sit a test. They won't forgive a surprise.
What does a realistic job preview actually include?
Four things, and most employers do one. Glider.ai's guide lists job shadowing, day-in-the-life video, sample work tasks or simulations, and candid Q&A with current staff about the job's downsides.
The ad is the first layer, and putting pay, shift pattern and the hard parts into it is covered in our frontline job description guide. Shadowing comes last, and trial shifts done right handles the legal side. This piece is about the middle: what happens between someone seeing your ad and someone booking an interview.
That middle is cheap to fix. A situational question costs nothing to write. A 30-second phone video costs an afternoon. A Q&A costs one honest employee and a coffee. Together they let the wrong candidates leave early.

How do situational questions double as a preview?
Write the question so that answering it means living the shift for a minute. Not "Are you comfortable with evenings?" but "It's 22:00, you're alone on the floor and two tables are complaining at once. What do you do first?" Anyone who reads that and thinks "absolutely not" has just previewed the job and screened themselves out. Anyone who answers well has shown you something a CV can't.
Three rules keep it honest:
- Describe demands, don't score them. Say the job means eight hours standing and lifting 15 kg crates. Don't ask candidates to rate their own fitness or availability, and never score anything that touches disability, age, gender or family status.
- One scenario per real hard part. If the worst bit is the Sunday close, the Sunday close goes in the quiz.
- Grade the reasoning, not the vocabulary. A good answer to the 22:00 question is about priority and calm, not polish.
This is how a Qwiza quiz works: the employer writes and grades the scenario, time per question sits alongside the score, and nobody is auto-rejected. A human decides.
An invented example. The numbers are illustrative, drawn from Qwiza demo campaigns and pilot targets, not client results. A 40-room hotel on the coast needs two night receptionists from December. The ad states 22:00 to 06:00, four nights on, three off, and the pay. Question three of the quiz is the 03:00 scenario: a guest locked out, the phone ringing, the fire panel showing a fault. What do you do first?
Of 120 people who open the quiz, around 70 finish. Roughly 30 stop at question three, which is the preview doing its job. Twelve are shortlisted, four interviewed, two hired. The previous winter the same hotel hired three from CVs and lost two inside a fortnight, both saying the nights were not what they had pictured.
What should a 30-second preview video show?
The hard part, filmed on a phone, without music. Employers instinctively film the highlights: the terrace at sunset, the smiling team, the new espresso machine. That's a brand ad. A preview video shows the 06:00 delivery, the walk-in freezer, the Saturday queue at 13:15. Let someone picture themselves there and decide.
One shift, one person, one voice. The best version is a current employee saying what a normal day looks like and the one thing they wish they'd known before starting. Format, length and running it as an ad are covered in video job ads. The preview-specific rule is simpler: if you'd be embarrassed for a new hire to see it on day one, it belongs in the video on day zero.
How do you run a candid Q&A without scaring everyone off?
Pick the employee who'd tell the truth anyway, and let them. Candid Q&A about the downsides is the format employers skip most because it feels risky. It isn't. The candidates who leave after hearing the downsides were going to leave in week two, when it costs more.
Three questions do most of the work: what's the hardest hour of the shift, what surprised you in your first month, and what do people underestimate about this job. Run it as a ten-minute call before the interview, a short clip on the careers page, or a paragraph in the ad under "What current staff say". Managers shouldn't answer. They'll soften it.
If night reception says the hardest part is being alone with nobody to call, that's a preview and a staffing problem in one sentence.
How do you know the preview is working?
Two numbers should move in opposite directions: more people leaving before the interview, fewer leaving after the start date. Aptitude Research (12-02-2026) found 26 per cent of organisations point to candidate drop-off during the hiring process as a challenge. Drop-off caused by an honest preview is the process working.
Track where people leave. At the situational question: the question did its job. After the offer: something between quiz and contract contradicted the preview, and the fixes for candidate ghosting apply. In week one: ask them, then put the answer into next month's ad.
Don't expect guarantees. The sources here show correlation and vendor claims, not controlled trials. What a preview reliably does is move the moment of truth earlier, from your payroll to their phone screen. For a Q4 seasonal hire, that's a quiet no in early November instead of an empty shift on 20 December.
Put the shift in the quiz
Qwiza turns one job description into an Instagram and Facebook quiz ad. Write the 22:00 scenario as a question, grade it yourself, and let the wrong fit leave early. First campaign free.
Try the live demoFrequently asked questions
Does a realistic job preview reduce the number of applications?
Yes, and that's the point. A preview that names the 06:00 start, the standing, the Sunday close and the pay will put some people off, and those are the people who would have quit in week one. In the Click Boarding survey of 2,000 US job changers (03-07-2026), 39 per cent of respondents named conflicting or misleading information as a reason for ghosting a new employer. Fewer, better-informed applicants cost less to interview and onboard than a full inbox of people who haven't understood the job. If applications fall to nothing, the problem is usually the pay or the rota, not the honesty.
Can I ask about physical demands or availability in a screening quiz?
Describe them, don't score them. It's fine, and sensible, to state that the job involves lifting 15 kg crates, eight hours on your feet or a 22:00 to 06:00 rota, so candidates can decide for themselves. It's a different thing to ask people to rate their fitness, list caring responsibilities or state their age, and then rank them on it. That path leads to disability, age, gender and family-status discrimination, and Meta's employment ad rules prohibit age and gender targeting anyway. Put the demands in the description, put the judgement in the scored questions, and keep a human on the final decision.
Does a realistic job preview replace a trial shift?
No. They do different jobs at different points. The preview works before the interview: it lets candidates who would hate the role discover that on their phone, at no cost to you. A trial shift comes after the interview, shows you how someone actually works, and needs to be paid, short and structured to stay fair and legal. Most frontline employers benefit from both. The preview thins the field honestly; the trial confirms the one or two people left. Skip the preview and your trial shift becomes the preview, which is the most expensive place to discover a mismatch.


