What a Bad Hire Costs, and Why a Salary Percentage Cannot Answer It

Why Doesn't a Percentage of Salary Work?

A percentage-of-salary figure travels from article to article without a document behind it, and the missing document is the smaller problem. A percentage assumes the cost of a hiring mistake scales with pay, and almost none of the four real inputs does. One multiplier cannot describe an exit at month four and an exit at month ten, and a multiplier tuned to either looks absurd on the other.

Take the two exits. A support hire who leaves at month four costs four months of a modest salary plus a fast backfill. A team lead who leaves at month ten costs ten months of a large salary, a long backfill, and everyone who spent those ten months reporting to someone the company was already unsure about. Nothing in the first scales into the second.

Then there is what such a figure includes. Almost every version folds in lost productivity, opportunity cost and morale, which are real and are also the three lines a CFO will ask you to substantiate first. When you cannot, the whole number goes with them. Keep the soft costs out of the headline figure and name them separately as things you believe but did not measure.

Build the Number From Four Lines You Can Source

Four lines, all internal, all defensible in a review: salary and benefits actually paid, recruiting spend you are about to repeat, ramp time bought and not delivered, and management hours consumed. Pull each one for the specific role family rather than for the company, because the four move in different directions across seniority and a blended average hides exactly the variation you are trying to price.

• Compensation paid to the exit. Loaded salary, employer taxes, benefits, any signing bonus not clawed back, plus severance if it applied. This is the only line finance already has to the dollar.
• Recruiting spend, repeated. Agency fees, job board spend, referral bonuses, travel, relocation, and the share of recruiter pay attributable to the req. SHRM's benchmarking puts median cost-per-hire at $1,244 for nonexecutive roles and $8,750 for executive roles, on a definition that covers those items plus advertising, job fairs and applicant tracking costs ¹. Read the whole distribution before you use the median: the same table shows a mean of $4,683 against that $1,244 median, so the spread is enormous and yours may sit anywhere in it.
• Ramp not delivered. Weeks of full pay before the role produced anything, doubled once, because you buy the ramp again on the backfill. Your onboarding plan almost certainly names a milestone you can use as the boundary, and how long a new hire takes to get up to speed is a number to read off your own last three hires rather than off a published ramp benchmark.
• Management hours. Coaching conversations, documentation, HR involvement, legal review, and the peer time absorbed by rework. Estimate it in hours at loaded rates and say out loud that it is an estimate.

The fifth thing to price is elapsed time, and it belongs to the backfill rather than to the exit. SHRM's median time-to-fill for nonexecutive roles was 44 days, with a quarter of organizations past 73 ². That is calendar time from opening the req to an accepted offer, so it excludes notice periods and the start date, and the vacancy is longer than the number looks.

What Does the Other Half of the Ledger Cost?

Roughly the same, and it produces no invoice, no exit interview and no post-mortem. A strong candidate your process rejected costs you the vacancy weeks you then spent finding someone else, the recruiting spend you paid twice, and whatever the eventual hire is worth less than they were. Nobody writes that down, because the rejected candidate never comes back with a performance review to prove the decision wrong.

That asymmetry is structural. Nobody is being careless. One error arrives with a name attached and a manager who remembers it. The other is a resume in an archive. Any process tuned only against the error that generates a story will keep tightening until it rejects on thin evidence, and the tightening will feel like rigor the whole way down.

Consider how good the underlying methods actually are. In the 2022 re-analysis of the selection literature, structured interviews came out top ranked at .42, ahead of job knowledge tests at .40, work samples at .33 and unstructured interviews at .19 ³. Those are corrected correlations with supervisor performance ratings rather than hit rates, and the .42 carries a wide credibility interval, but the direction is unambiguous: the best evidenced single method in the field still leaves a great deal of both errors on the table. A process that never misses a good candidate does not exist to be bought.

So the honest business case has two numbers in it and a stated ratio between them. Which mistake your process is actually built to avoid is the question that ratio answers, and it is a different question from how much either mistake costs.

Write the Number Down Before You Need It

Do it once, for one role family, while nothing is on the line. A figure assembled during a budget argument reads as motivated, and the person you are asking will treat it that way. A figure assembled last quarter, with the four lines shown and the assumptions labelled, reads as a measurement someone bothered to take.

The worked version fits on one page. Name the role family and the level. Show the four lines with their sources. State the two estimates you could not source, in a separate block, with the range you would accept. Then add the count: how many of these happened last year, at what rate, and what a percentage point of change in that rate would be worth. The rate is what any process change actually moves, and a per-incident cost without a rate cannot be turned into a budget.

One caution about attribution. Most departures inside a year are not selection failures at all, and lumping them together inflates the number in a way that is easy to attack. Separate the ones where the evidence you gathered pointed the wrong way from the ones where the role changed, the manager left, or the offer was accepted reluctantly. Only the first group tells you anything about the hiring process, and only that group belongs in a case for changing it.

Then keep the file. When the process changes, you will want to know whether the rate moved, and proving a hiring change actually worked starts with having written down the before. Pair it with a definition of quality of hire the hiring manager does not own, and the same page answers both the cost question and the improvement question. For the specific failure the finance team will ask about first, the candidate who interviewed brilliantly and struggled by month three is the case worth reading.

See a sample report >