The hidden cost of a bad hire

Search this and you will be told a bad hire costs 30% of first-year salary, or 3× salary, or £132,000. Those figures come from surveys in other countries with other labour laws, and they disagree with each other by an order of magnitude. We are not going to add a seventh number. We are going to show you which parts of it you can work out from records you already have, and which parts nobody can.

Why the famous numbers don't transfer

The multipliers you find are not fabricated, but they are answering a different question than yours. Three reasons they do not survive the journey:

• Notice and severance are local. What it costs to end an employment relationship is set by law and by contract, and those differ by country. A figure built on three months' notice does not describe a role with one.
• Every study counts different things. Some count only recruitment and salary paid. Some add lost productivity. Some add the manager's time, the team's time, customer impact, and a guess at morale. The multiplier moves enormously depending on where the author stopped.
• They average across roles that have nothing in common. A bad hire in a two-person finance team and a bad hire in a fifty-person call centre are not the same event, and averaging them produces a number that describes neither.

The four things it actually costs you

Underneath the multipliers there are four real components. Two you can price from your own records this afternoon. Two you cannot, and should stop pretending you can.

1. Doing the recruitment again — knowable

The whole of your original hiring cost, repeated. Job posting, the hours spent reading and interviewing, any agency fee. We have broken that down line by line: the posting has a published price, the processing is measurable, and your team's time is the part only you can price. Whatever that came to the first time, it is due again.

2. The empty seat — knowable

From the day they leave to the day a replacement is productive, that work is either not happening or is being absorbed by people who already had jobs. You know roughly what that role was supposed to produce, and you know how long your last hire took to fill. Multiply.

This is usually the largest of the four, and almost nobody counts it, because it does not appear on any invoice.

3. Ramp-up paid twice — partly knowable

Nobody is fully productive on day one. If it takes three months to get there, and your bad hire left at month five, you paid for three months of ramp and got two months of output — and now you will pay for three months of ramp again.

You can estimate your ramp period honestly by asking the manager when the last person started carrying their own work. That is not a precise number, but it is your number, and it beats a multiplier from a foreign survey.

4. What it does to everyone else — not knowable

A bad hire in a small team means the others cover, get frustrated, and sometimes leave. This is real. It is also not measurable in any way we would defend, and the studies that put a figure on it are the least reliable part of the literature. Name it in your reasoning, refuse to price it, and be suspicious of anyone who does.

A self-audit you can actually do

This takes about an hour with your own records and produces a number that is true about your company rather than about a survey panel:

1. List every person who left within twelve months of starting, over the last two or three years. Voluntary or not.
2. For each, write down three dates: when they started, when they left, when the replacement started carrying the work.
3. Count the weeks the seat was effectively empty — from their departure to the replacement being productive, not to the replacement's first day.
4. Add your recruitment cost, twice. Once for the original hire, once for the replacement.
5. Divide by the number of hires you made in that period. That is your real cost of hiring badly, spread across all your hiring.

Two things usually surprise people. The empty-seat weeks are longer than remembered — the gap between "replacement started" and "replacement productive" is where the cost hides. And the rate is either much lower or much higher than the industry figures, because your hiring is not an industry average.

What the number is for

Not for a slide. For one decision: how much evidence is it worth gathering before you make an offer?

If a wrong hire costs your company the equivalent of four months of that salary, then an extra week of process — a work sample, a second conversation, a reference call that is actually made — is cheap insurance. If it costs three weeks, the extra process may be the more expensive mistake. The answer is different for a call-centre role and a finance lead, which is exactly why one industry multiplier cannot serve both.

What reduces it

• Requirements you would actually reject over. Most bad hires clear a list that was never really the job. Three to five real requirements beats a wall of nine tools.
• Evidence instead of impressions. Anything that makes the candidate demonstrate the work rather than describe it.
• Reading the whole pile. A rushed shortlist is the most common origin of a mediocre hire, and stopping halfway is how people rush it.
• Consistent comparison. Candidate number 40 judged by the same standard as candidate number 4.

What does not reduce it

Adding interview rounds without adding information. A fifth conversation with the same three people asking the same questions feels like diligence and produces nothing a second conversation did not. Rigour is about what you learn, not how many meetings you hold.

And no scoring system prevents a bad hire, including ours. A CV score reads a document; it cannot tell you whether someone will be difficult to work with. It narrows who you spend your evidence-gathering on. The evidence-gathering is still yours to do.