By the time a bad hire is visible, the mistake is usually months old: the role was never defined, so the ad attracted everyone, the interviews measured charm against vagueness, and the new hire failed at expectations nobody had written down. The antidote costs one page and one hour before anything is advertised - and it upgrades every step that follows automatically.
- Mis-hires are usually mis-definitions: the role was a vague wish before it was a bad choice.
- Define outputs (what must be true because this person exists), then derive tasks, then requirements - most people work backwards and inflate.
- Three must-haves maximum; everything else is trainable or preference. Every extra must-have deletes good applicants.
- The 90-day success sentence - written before advertising - becomes your screening criteria, your onboarding plan and your probation review.
Outputs first: the existence test
Start from the question that cuts through: what must be true in three months because this person exists? Not tasks - outcomes: 'the weekend rush runs at current review scores without me on the floor'; 'every order ships same-day again'; 'the phone gets answered and the diary stays full'. Three to five such sentences are the role. Tasks derive from them, requirements derive from the tasks - defining in this order prevents the universal inflation that starts from 'what would the ideal person be like' and ends demanding a unicorn for a counter job, the trap behind half of the classic hiring mistakes.
The must-have discipline
- Three maximum, each passing the rejection test: you would truly turn down an otherwise excellent candidate over it.
- Availability is usually one of them - the shifts are the job, so they gate honestly, per screening questions.
- Name the trainables out loud: 'we teach the till, the menu and the machine' widens the pool exactly where skills-based hiring says the talent hides.
- Kill the proxies: years-of-experience and degree lines almost always stand in for capabilities you can screen directly - screen the capability instead.
The honesty paragraph
Add a day-in-the-life paragraph written for truth, including the hard parts: the 6am starts, the lifting, the difficult-customer quota, the repetition. This is the text that becomes block two of the ad, and its honesty is load-bearing: candidates who accept the real job stay past week three; candidates recruited by omission become early-turnover statistics with your training investment in their pocket.
The 90-day sentence, and what it automates
Finish with one sentence: 'This hire has succeeded if, by day 90, ___.' Write it before advertising and it becomes, unchanged: the screening scorecard (questions that predict exactly that sentence), the interview structure, the onboarding plan (whatever the sentence needs, week one teaches), and the probation review - held against criteria the candidate heard on day one, not invented at month three. One page, one hour, and the whole hiring chain inherits its clarity.
The takeaway
Define outputs, derive tasks, cap must-haves at three, write the honest day and the 90-day sentence - one page before any ad. Most hiring quality is decided in this hour; everything afterwards is execution of whatever you wrote, or of whatever you failed to.
From definition to campaign.
Give Qwiza the role and the must-haves - it builds the ad and the scored screening around exactly them. Ranked candidates against your criteria, 48-hour pilot target.
See how Qwiza worksFrequently asked questions
What belongs on a one-page role definition?
Five blocks: the outputs (three to five things that must be true because this person exists - 'weekday lunch service runs without the owner'), the recurring tasks that produce them, the genuine must-haves (three maximum: the licence, the availability, the non-trainable capability), the explicit trainables (what you will teach, which widens the pool honestly), and the 90-day success sentence. One page forces the prioritisation that forty-line job descriptions exist to avoid.
How do I know if a requirement is a real must-have?
Two tests: would you genuinely reject an otherwise excellent candidate lacking it (if not, it is a preference), and is it learnable in your first month at acceptable cost (if yes, it is a trainable). Years-of-experience requirements fail both tests most often - they proxy for skills you can measure directly, per skills-based hiring, and every inflated must-have deletes exactly the returners, career-changers and young talent the market says you need.
Who should be involved in defining the role?
Whoever the hire will actually work beside, for fifteen minutes: the current shift lead knows the real tasks and the real gaps better than the org chart does. If the definition is contested - two owners wanting different roles from one hire - settle it before advertising; the classic disaster is interviewing candidates against two secret, incompatible scorecards, then blaming the hire for failing one of them.
