The True Cost of a Bad Technical Hire

The number that gets passed around is 30 percent of first-year salary — that's the U.S. Department of Labor's old estimate for the cost of a bad hire. After twenty years in staffing, we'll say plainly: for technical roles, that number is off by a multiple. The recruiting fee and the salary are the cheap part. Here's the honest accounting.

The visible costs are the small ones
Salary, benefits, the recruiting cost, onboarding time — for a $120,000 engineer who lasts eight months, call it $100,000 all-in. Painful, budgetable, and not the real story.

The invisible costs are where it gets expensive
Bad code compounds. A struggling engineer doesn't produce nothing — they produce work that other people have to review harder, fix later, and build around. One weak senior engineer can quietly convert a healthy team into a team that spends a third of its time on remediation. Months after the person is gone, their decisions are still in the codebase, still charging interest.

Strong people carry the load, and they keep score. When someone isn't pulling their weight, the work doesn't wait — it lands on the two or three people you can least afford to burn out. They'll do it, for a while. But every month a manager doesn't act, those people are updating a private ledger about whether leadership notices anything. The most expensive outcome of a bad hire is a good resignation, and we've watched that exact sequence more times than we'd like.

Managers lose their most productive hours. A struggling hire consumes performance conversations, documentation, re-planning, and the slow-motion agony of a formal process. Six months of that attention, aimed at your best people instead, would have compounded. Aimed at a failing hire, it just slows the bleeding.

And the seat is still empty. At the end of it all, you're back where you started — except the team is tired, the project is late, and the search clock restarts from zero. A bad hire isn't an alternative to a vacancy. It's a vacancy with an eighteen-month delay and interest.

Where bad technical hires actually come from
Rarely from lying candidates. In our experience the source is almost always the process: interviews that tested trivia instead of the actual job, nobody checking whether the person had done the specific kind of work (building from scratch versus maintaining at scale — different animals), culture-fit conversations that mistook charm for collaboration, and reference checks skipped because everyone was in a hurry. Speed pressure is the classic trigger — which is bitterly funny, because the rushed hire costs a hundred times more than the extra two weeks of process would have.

This is also the honest case for screening depth in staffing. When a firm technically screens before submitting, checks the actual work history against the actual need, and has placement history on the person — the interview-to-hire ratio tightens and the failure rate drops. That's the product, far more than the resume pile is.

Common questions
How fast can you tell a hire has gone wrong?
The team usually knows by week six; the manager admits it around month four; the org acts around month eight. Shrinking that middle gap is worth more than almost any other hiring process improvement.

Is contract-to-hire the answer?
For genuinely uncertain cases, yes — evaluating real work beats any interview. But it's a tool for uncertainty, not a substitute for screening. A bad contractor still costs you months.