Sample frameDay 244·Hiring · Select

Good morning. There's a ranked list of every way to judge a candidate. Your process is built from the wrong end of it.

Hiring.

The Predictor Ranking

5 min read·Apply by lunch

The question

Add up the hours your last hire's process actually consumed — the CV sift, the coffee, the panel conversation about whether they'd fit — and then add up the hours that produced a written score against a stated outcome: which number is larger, and what does that ratio say about what you truly believe predicts performance?

The idea

In 1998 Frank Schmidt of the University of Iowa and John Hunter of Michigan State published a paper in Psychological Bulletin with an unusually literal title: the validity and utility of selection methods, drawn from eighty-five years of research findings. They'd pooled thousands of studies to answer one question — for each way of assessing a candidate, how well does the assessment predict how that person later performs? The output is a ranked list, and it remains the most useful single document in hiring. Work samples, tests of general mental ability and structured interviews sit at the top. Unstructured interviews sit well below them. Years of job experience and years of education sit lower still. Handwriting analysis scores roughly zero, which is the number an honest process should also give to a gut feel. Paul Sackett and colleagues revisited the arithmetic in the Journal of Applied Psychology in 2022, arguing the older figures had been over-corrected for range restriction; their revised list moves the numbers about — structured interviews come out on top, cognitive tests lower — but the shape holds. Samples and structure beat impressions and credentials. The trap is that the two instruments most hiring actually runs on, the CV and the chat, are the two the evidence ranks near the bottom, and they're the two that feel most like knowing someone.

The framework

01
Count the minutes — the real ones, out of the diary.
Take the last role you filled and walk it stage by stage, putting a number against each one: the sift, every conversation, the panel, whatever scoring happened, per candidate. Do it from the calendar rather than from memory, because memory reports the stages that felt significant while the diary reports the stages that consumed your week. Most people can't produce these numbers at all, which is itself the finding — you can't have been managing an allocation you've never once measured. The whole of this morning is that single column of minutes, set beside the list above.
02
The list is a bet across many hires, not a verdict on one.
Every number on that ranking is an average pooled across thousands of studies, which means none of it tells you anything about the person sitting in front of you this morning. What it tells you is which instruments pay off if you run them fifty times. That distinction matters, because every experienced manager owns a story about the brilliant hire a rigorous process would have rejected — and one such story is precisely what a correlation of a half predicts you'll have. Read the ranking as policy for a portfolio of hires. It was never an argument about an individual candidate.
03
The ceiling nobody quotes.
Even the strongest instruments correlate with later performance at around a half at the optimistic end, which accounts for perhaps a quarter of what happens next. So a good ranking is a way of being wrong less often, not a way of being right. The practical consequence is restraint: keep the process cheap enough that you'd happily run it again, treat the first ninety days as the real test, and resist building a nine-stage gauntlet in pursuit of a certainty the evidence says isn't on offer. Length is not rigour. It's usually anxiety, wearing rigour's clothes.
04
Where your hours sit is what you actually believe.
Finish the audit with two lines written next to each other: the instrument you'd name, in the abstract, as the best predictor of performance, and the instrument your diary actually funds. For most processes those are different sentences, and the reason isn't ignorance. The weak instruments are the cheap ones, the fast ones and the ones that feel like getting to know someone, so they win the calendar without ever winning the argument. Hold on to that gap. The mornings after this one take the instruments this elective actually runs — the screen, the interview, the work sample, the reference call — one at a time, and show you what each costs to run properly. The ability test isn't among them, and that's deliberate: at your size a validated one you could defend is a purchase and a legal opinion rather than a morning's work, and a work sample reaches most of what it would have told you.

Members only · 4 principles + template + AI mentor

364 more frameworks are waiting.

The Predictor Ranking is Day 1 of 365. One framework every morning for a year — across strategy, sales, negotiation, leadership, and more.

$1 /day

Billed $365/yr · cancel any time

365 frameworks

12 topics · templates · AI mentor

Reserve your place — $365/yr →

One payment a year · the whole library · no auto-renewal.

More Hiring frameworks

All 19 hiring frameworks →