Hiring a payroll administrator
Eighty applications for a role where being right ninety-nine per cent of the time means somebody did not get paid this month — and where the whole job compresses into three days at the end of it.
Nobody puts their mistakes on a CV, and mistakes are the entire risk. Payroll is unusual in that the failure mode is specific, immediate and personal: a wrong figure is somebody’s rent, and it arrives with a deadline that legally cannot move. What predicts a good payroll hire is therefore not breadth of experience but temperament under a fixed deadline and a habit of checking — and neither is visible on a document that lists software packages and headcounts processed. Statutory knowledge is the other half, and it dates: someone who has not run payroll for two years has missed real legislative change. The classic false positive is a long payroll history in a single stable company where nothing unusual ever happened.
Headcount, complexity, and what happens at month end
Selene goes after the shape of the payroll run, because a simple monthly salary run and a weekly mixed-rate one are barely the same job.
Frequency, variation, and who checks the work
Payroll complexity comes from variation rather than headcount: shift premiums, weekly and monthly runs together, high turnover, salary sacrifice. Establishing that is more useful than asking how many employees.
One sentence changes the guide
It also names why payroll errors cluster where they do. Almost nobody makes mistakes in week one of the cycle; they happen on the last afternoon, under time pressure, on the run that has already reconciled once. A candidate who describes a deliberate final check has told you something no accuracy claim can.
Exceptional, not just good —
“The one who checks it again on the last day, when they are tired and it has already balanced.”
→ A described checking routine became a must-have; years in payroll dropped below it.
What the guide ends up measuring
Six criteria, three of which are about behaviour rather than knowledge. Software sits at weight 1 deliberately: every applicant lists a package, the packages are learnable in a week, and none of them prevents the error that actually costs you.
Two facts, and a currency question that matters more than it looks
Right to work is a straightforward gate. The second question — when they last ran a live payroll — is not about eligibility so much as about decay: statutory rules change every April, and a candidate who last ran payroll three years ago is not less capable but is definitely out of date, which is a different conversation and worth having on day one rather than in month two.
Scored on the habit, not the headcount
Selene never receives the candidate’s name or the raw CV. She scores a blind, structured profile, with any stated age, sex, nationality and religion stripped out before scoring runs.
She reads for described routines and caught errors, because a claim of accuracy is the one thing every application in this pile makes and none of them evidences.
How the blind scoring works →What stays with you
Selene's half
- Establishing the real complexity — variation, not headcount — before writing criteria
- Reading all 80 against the same guide, blind, with the evidence attached
- A ranked shortlist built on described habits rather than accuracy claims
- An interview brief aimed at the statutory areas each candidate did not mention
Yours
- Advertising the role — Selene doesn’t post to job boards or source candidates
- Any payroll test or scenario, which on this role is worth doing properly
- The interview, with her brief in hand
- The offer, and every call that matters
Nearby roles
Describe the variation
Weekly and monthly, overtime, shift premiums — complexity is the hire here, not headcount.