Cost of a Bad Hire: The Average and True Cost of Hiring a Bad Employee, and Where the Numbers Come From

A bad hire costs far more than the salary you paid them, and almost every number you will read for it comes from a survey rather than a measurement. Built from data that does have a traceable source, a $100,000 US hire who leaves in month five costs roughly $36,000 to $39,000 if you recruited direct, and about $77,000 if you paid a 20% agency fee on the failed search and again on the replacement.

That first figure works out to 37% of first-year salary, which lands close to the 30% rule of thumb everyone quotes. The difference is that you can rebuild it from your own payroll and see which line is actually hurting you. Here is the arithmetic, and first, an honest look at where the famous statistics came from.

Where does the 30% figure come from?

If you have read three articles on this topic you have read the same two claims: that the US Department of Labor puts the cost of a bad hire at 30% of the employee's first-year earnings, and that CareerBuilder found companies lose an average of $14,900 per bad hire. Neither one usually carries a link.

We went looking. Here is what each claim actually rests on.

None of that means the numbers are wrong. The 30% figure is probably conservative, as our own build-up below suggests. It means you should not put a statistic in a budget request when you cannot open the study behind it, because the first competent person who asks where it came from will find what we found.

The good news is you do not need it. Every input for a real calculation is either sitting in your payroll system or published by an agency that does show its work.

How much does a bad hire cost? The build-up

Start with what the person cost you while they were there. That is not their salary, it is their fully loaded cost: salary, plus employer payroll taxes at 2026 statutory rates, plus the employer share of benefits.

For a $100,000 salaried US hire that comes to about $131,400 a year, or $10,947 a month. The components are 6.2% Social Security on the first $184,500 of wages, 1.45% Medicare with no cap, 0.6% FUTA plus state unemployment, and roughly 23% in cash benefits derived from the Bureau of Labor Statistics March 2026 compensation data. The full derivation, including why most calculators overstate it, is on our page for the cost of hiring an employee.

Now assume this hire lasted five months and delivered about half the output a fully ramped person would have. That is a deliberately generous assumption, because a hire who is being managed out usually delivers less than half in their last two months.

The direct-hire total is 37% of first-year salary, and the agency version is 77%. Same person, same five months, same mistake. The single line that doubles the damage is the placement fee, because you pay it on the search that failed and you pay it again on the one that fixes it. That asymmetry is why the agency question deserves more thought before a risky hire than after one, and it is costed model by model on our recruiter fees page.

The time figures come from published benchmarks rather than estimates. SHRM puts median time to fill at 39 days for nonexecutive roles, and technical roles run considerably longer, which we cover in average time to hire a software engineer. Two and a half months of an empty seat is a real cost even though nobody invoices you for it.

What are the hidden costs of a bad hire?

The table above is the part you can defend in a budget meeting. The part that usually hurts more never reaches an invoice.

Rework. Bad hires rarely produce nothing. They produce work that someone else has to check, correct or redo, which means the cost lands on your strongest people, who are the most expensive and the most likely to resent it. In customer-facing seats the damage lands outside the company entirely, in the support and back-office operations that absorb the mistakes and in accounts that quietly stop renewing without ever telling you why.

Manager attention. A struggling hire consumes a wildly disproportionate share of a manager's week: extra check-ins, documentation, HR conversations, and the mental overhead of a decision they keep postponing. That time comes directly out of managing everyone else.

Team morale and the second departure. This is the expensive one. Good people notice when someone is not carrying their weight and notice again when nothing happens about it for four months. The cost of a bad hire sometimes shows up as a resignation from a completely different person, and by then nobody connects the two events. What keeps that from happening is covered in remote employee retention.

The delay before you admit it. Most managers know within about six weeks and act at around five months. Those intervening months are almost pure loss, and they are the only line on the whole list you can shorten for free.

How do you reduce the cost of a bad hire?

Four things move the number more than anything you can do in an interview room.

Publish the salary range. It is the cheapest filter that exists. Candidates who would have quit over money within six months screen themselves out before they cost you an interview slot, and the ones who apply anyway have already accepted the offer in principle. Against the 254 to 300 plus applications that Greenhouse and Ashby now measure on open roles, a filter that works before you read anything is worth more than any assessment you run afterward.

Replace the fifth interview with a paid work sample. Conversational interviews measure how well someone interviews. A short paid exercise resembling the actual job measures the actual job. It also gives the candidate a real preview, which prevents the failure mode where a good hire leaves because the work was not what they were sold. Our guide to interviewing remote candidates covers how to structure one without burning a week of your team's time.

Run a real first week. A meaningful share of hires labelled bad were fine hires who were left to figure it out alone, produced nothing visible for two months, and were judged on that. The fix is unglamorous and cheap: a written plan, an owner, and a first task that ships. The remote onboarding checklist is the artifact.

Keep the recruiting cost small so failure is survivable. This is the structural one. If a failed hire costs you $3,400 to redo, you will act on the evidence in month two. If it costs you $20,000, you will find reasons to wait, and the waiting is what makes it expensive. For roles where qualified people simply do not know the job exists, which is most professional roles and nearly all remote ones, distribution is the whole job: a listing on a remote-only board is $299 with the salary range required, and there is no percentage riding on whether it works out. The comparison across channels is in cost to post a job, and the full budgeting picture is on cost of hiring an employee.

The number worth writing down

If you take one figure from this, do not take 30%. Take the method: fully loaded monthly cost, times months paid, times the share of output you did not get, plus both recruiting costs, plus the second vacancy. Run it once on a hire that actually failed at your company and you will get a number that is specific, defensible, and almost certainly larger than the one you have been quoting.

You will also see immediately which line is doing the damage, and it is usually one of two: the months you waited before deciding, or the fee you paid twice. Both are within your control, which is more than can be said for most costs on a P and L.