The seat was filled. That was the problem.
Four months of interviews. Two rejected offers. A dozen arguments with your VP of Engineering about whether the candidate was "senior enough." Then, finally, a signed offer letter and a start date. The Senior Backend Engineer showed up Monday, and you exhaled. The req was closed. You could stop recruiting and start shipping.
By day 90 the exhale has turned into a knot. They're still asking the same questions they asked in onboarding week. Two of your strongest engineers keep "pairing" with them — which is a polite way of saying redoing their pull requests. You spent last Sunday night rewriting the architecture doc they were supposed to own. You have a mis-hire, and some part of you knew it by week three.
Here's where most Series B and C founders get the math wrong. You paid them $160,000 base. You let them go at six months. So you write the loss down as roughly $80,000 in salary and call it expensive but recoverable. That number is off — not by a little, by an order of magnitude. In a company where shipping velocity is the thing investors are actually pricing, a failed engineering hire isn't a sunk cost sitting still on the P&L. It's an active drag that pulls down the people around it every single day they're employed and for months after they're gone.
Dr. Bradford Smart, who wrote Topgrading, pegs the true cost of a mis-hire at 5x to 27x base salary depending on seniority. At your stage that's not academic. For an individual contributor the real number rarely lands under $240,000, and for a VP-level miss it clears seven figures fast. The trouble is your accounting system can't see most of it. So let's do the math your P&L refuses to show you.
Four buckets, and only one of them shows up in QuickBooks
Take a real shape: a Senior Engineer hired at $160K base, terminated at month six. Run a forensic audit and the damage sorts into four buckets. Your CFO can see the first one. The other three are where the company actually bleeds.
Bucket one — the cash that left the building
This is the only bucket your finance team will recognize on its own:
- Fully loaded salary and benefits, six months: ~$100,000
- Agency fee at 20% of base, non-refundable past the 90-day guarantee: ~$32,000
- One month severance: ~$13,000
- Legal and admin to exit cleanly: ~$5,000
That's roughly $150,000 in hard cash for a person who shipped nothing you kept. Painful, survivable, and — critically — the small bucket.
Bucket two — the tax your best people pay
A mis-hire doesn't just produce zero. They consume the output of the people who have to compensate for them. The SHRM data on mis-hires puts it bluntly: managers burn 17% of their time managing poor performers. So if your CTO loses one day a week to unblocking, re-reviewing, and re-explaining, you are routing 20% of your most expensive technical leader's salary into subsidizing a failure. That's not the mis-hire's cost. That's the cost of the talent you were trying to protect.
Bucket three — the feature that never shipped
This is the bucket that should keep you up at night. You hired this person to deliver something specific — say, the multi-tenant billing rework that unlocks your next enterprise tier. That work is now delayed by the six months they were here, plus three months to find a replacement, plus three months for that replacement to ramp. Call it a twelve-month slip on the roadmap line you were counting on. If that line was modeled to add $1M of ARR in its first year, the mis-hire didn't cost you $150K — at a 6x revenue multiple, it quietly erased $6M of enterprise value from the number you'll defend in your next raise.
Bucket four — the code that has to be deleted
In engineering there's a category most cost models miss: negative work. A weak hire doesn't leave behind nothing — they leave behind commits that have to be unwound. Two sprints of your two best engineers refactoring brittle, undocumented code is two of your highest-leverage people producing no new value for a month. And it metastasizes: LinkedIn Talent Solutions turnover data ties the majority of regretted attrition back to hiring decisions, not compensation. Your A-players will tolerate carrying a C-player for a quarter. They will not tolerate it for a year — and the ones you can least afford to lose are exactly the ones with the strongest outside options. The U.S. Department of Labor cost-of-turnover framing is the floor here; the compounding you can't expense is the real liability.
Stop trusting your gut. Build the gate instead.
You can't afford a second $240,000 lesson, and "interview harder" isn't a system. The fix is to replace intuition with three pieces of infrastructure you can install before your next offer goes out.
Write a scorecard, not a job description
Most mis-hires trace back to a vague req. "Must be a self-starter who thrives in ambiguity" is a horoscope, not a hiring bar. Replace it with the outcome the seat exists to produce: "Within 90 days, this engineer ships the billing-API rework for our top two enterprise accounts with zero critical defects and full test coverage." If you can't write the win down in one sentence, you have no way to recognize the person who can deliver it — and no way to know, at day 90, whether they did.
Make the technical screen mirror the actual job
The most expensive three words in hiring are "I liked them." Energy in a final-round conversation predicts nothing about whether someone can navigate your codebase under deadline pressure. Run a work-sample assessment that looks like a real Tuesday on your stack, not an algorithm puzzle off a whiteboard. If you're hiring a manager, the assessment is different but no softer: pull their last team's retention. You are buying judgment and the ability to keep good people, not raw coding speed.
Install a 30-60-90 with a real off-ramp
The only thing more expensive than a mis-hire is a mis-hire you keep for a year because firing feels worse than waiting. Set hard checkpoints. Missing the 30-day signals? Intervene directly, now. Still missing at 60? Start preparing the exit — don't open a six-month improvement plan when you already know how it ends. The pattern in post-acquisition attrition benchmarks holds here too: decisive, early breaks protect team morale, while drawn-out limbo corrodes it and pushes your A-players toward the door.
Do this Monday: pull your last failed hire and run the four buckets — cash out, manager drag, the slipped roadmap line at your revenue multiple, and the cleanup sprints. Put a single number on it. Then write the one-sentence scorecard for the next person who fills that seat. Your valuation is a function of your team's ability to execute. Every mis-hire is a leak in that function — and now you can see exactly how big it is.
