When a hire doesn't work out, most founders go straight to the number they can point to — the salary paid out before anyone admitted it wasn't working. That number is real, but it's also the easiest one to calculate, which is usually a sign it isn't the whole story.
The actual cost of a bad hire is spread across months, buried in places that don't show up on a single invoice. By the time it's obvious enough to act on, most of the damage is already done.
The obvious cost is the one you already priced in
Salary, payroll tax, a laptop, maybe a signing bonus — you budgeted for all of it before day one. Losing that spend hurts, but it's a known quantity. Founders rarely regret the number on the offer letter. They regret everything that happened around it.
The ramp-up you paid for twice
Nobody is productive in week one, and that's fine — it's expected. What isn't priced in is that when the hire doesn't work out, you pay for that same ramp-up again with whoever comes next. Two onboarding periods, two rounds of "let me show you how we do this," and only one of them produced anything.
Your own time is the real line item
This is the cost nobody puts in the spreadsheet. Every week a hire is underperforming, someone has to notice, coach, redirect, and eventually decide — and that someone is usually you. Managing a mismatch quietly consumes the same hours you hired the role to free up in the first place. You end up doing the job and managing the person who was supposed to be doing it.
Why the decision gets delayed
Everyone knows, in theory, that a bad fit should be addressed quickly. In practice it rarely is. There's sunk cost — you already spent weeks hiring and training this person. There's hope — maybe it's a rough patch, not a mismatch. And there's the sheer effort of starting the search over. All three push toward waiting, and waiting is exactly what turns a bad hire from expensive into very expensive.
What actually lowers the risk
None of this is an argument against hiring — it's an argument for shortening the distance between "this isn't working" and doing something about it. That's easier when the commitment on both sides is smaller to begin with. A trial period before anything is signed doesn't just reduce the visible risk; it removes the sunk cost and the hope that keep founders stuck managing a mismatch for months longer than they should.