The bad hire — a CFO’s model.

If you’re a CFO, you’ve seen the headline number: a bad hire costs the equivalent of 30% of the role’s annual salary. Maybe more. The number is roughly correct, but it’s also useless without the breakdown. You can’t make a budgeting decision on a single composite metric. So here’s the model, broken down by line item, with the assumptions made explicit.

The four cost categories

Every bad hire generates costs in four buckets. The dollar weights vary by role, level, and industry, but the structure holds.

1. Recruiting cost — sunk

What you already spent to source, screen, and onboard the hire. This is unrecoverable.

• Recruiter time: 23 hours × loaded recruiter cost. ($1,610 US, ₹34,500 India)
• Hiring manager time: 8–12 hours of interviews and panel reviews. ($1,200–$1,800 at typical engineering manager rates)
• External agency fees, if any: 15–25% of first-year salary, paid up front and typically non-refundable past 90 days
• Background check, equipment provisioning, onboarding cost: $500–$2,000 depending on remote/in-office

Subtotal for a mid-level US engineer: $3,500–$8,000 per bad hire, before agency fees. With agency: easily $25,000+.

2. Ramp & productivity loss

The bad hire is on your payroll while not contributing meaningfully. Typically 3–9 months before you and they admit the mismatch.

• Salary paid during ramp: 6 months × salary. (For a $120K engineer: $60,000)
• Reduced output from the team supporting them: 15–25% productivity tax on 2–3 colleagues for the same duration
• Manager time spent on performance management: 3–5 hours/week for 2–3 months before the decision to part ways

Subtotal for a mid-level engineer: $75,000–$110,000 in salary plus productivity drag.

3. Exit cost

Severance, replacement recruiting, knowledge transfer (or, more often, knowledge loss).

• Severance: 2–8 weeks of salary, depending on jurisdiction and policy
• Replacement recruiting: roughly the same as the original recruiting cost — the role is open again
• Knowledge transfer: usually negligible (the bad hire didn’t accumulate transferable institutional knowledge)
• Team morale cost: hard to quantify, but real. Typically 1–2 voluntary departures triggered indirectly by a notable bad hire

Subtotal: $15,000–$30,000 direct exit cost, plus the next round of recruiting.

4. Opportunity cost

The biggest line item, and the one most often missed. Every month a role is filled by the wrong person is a month not filled by the right person.

• Delayed feature shipment, missed deal, deferred roadmap progress
• Customer churn linked to under-delivery on the bad hire’s scope
• Strategic project deprioritised because the team that should have done it is busy compensating

This is the unbounded one. For an executive bad hire (VP, C-level), opportunity cost can dwarf all other categories combined.

The total, fairly modeled

For a mid-level US engineering hire at $120K salary, with an industry-average 6-month bad-hire lifecycle:

• Recruiting cost (sunk): $5,000
• Ramp & productivity: $90,000
• Exit cost: $20,000
• Opportunity cost: $30,000–$200,000 (highly variable)

Total range: $145,000 to $315,000 per bad hire. The often-quoted “30% of annual salary” number ($36K) is roughly the visible direct cost — recruiting plus exit. The other 80% of the true cost is below the waterline.

The economic question for AI hiring isn’t “does it pay for itself.” It’s “how many bad hires per year does it prevent.” Even one prevented bad hire on a 100-hire-per-year operation pays for the platform many times over.

What to ask your TA team

1. What was our bad-hire rate (people who left or were exited within 12 months) over the last three years?
2. What did each of those cost us, using the framework above?
3. What would a 30% reduction in bad-hire rate be worth, in dollars, annually?
4. Is our current screening process specifically designed to reduce that rate, or is it designed to reduce time-to-hire?

If the answer to (4) is “time-to-hire,” you have your answer about where to invest first.