One Bad Warehouse Hire Burns $22K in 90 Days: Probation Math and the Legal Line

A bad hire doesn't cost you their wages. It costs you the slot they occupy — for 90 days, that slot produces at half the rate of a normal worker while eating double the management attention.

I once ran the numbers for a 3PL warehouse in California's Inland Empire: one wrong hire at $20 an hour, a regular picker, burns well over $15,000 in sunk cost across 90 days. And that's the clean version — the person leaves quietly at day 90 and nobody sues anybody. Drag it out to day 120, or leave a discrimination claim lying around, and you can add another zero.

What one wrong hire actually burns

Line by line, using common West Coast warehouse assumptions (illustrative estimates based on a $20/hour picker; adjust to your local reality):

Total: roughly $22,000 for one wrong hire over 90 days. Again — that's the amicable version, where nobody lawyers up.

Here's the counterintuitive part: most people treat probation as a cushion, a 90-day window to "wait and see." Wrong. Probation is a stop-loss window, not an observation window. If you can see the problem at day 30, waiting until day 90 to act just burns 60 more days of full wages and one slot's worth of low output. Cutting early isn't ruthless — it's arithmetic.

That Inland Empire warehouse changed its rule: a hard review at day 30. Miss the bar, get a direct conversation and a 2-week improvement plan. Still missing, you're gone. Within a year their probation termination rate rose from 8% to 15% — and total labor cost fell 11%, because slots turned over faster and bad fits stopped collecting full paychecks.

The three measurable lines of probation

"He just doesn't feel right" is worth nothing in court and nothing on a P&L. Probation has to be quantified — three lines, every one measurable:

Line one: throughput. By day 30, UPH (units picked per hour) hits 60% of the seasoned-worker average. Day 60: 80%. Day 90: 95%. Miss it and you miss it. There's no "but his attitude is great" buffer clause.

Line two: quality. Error rate is tracked separately, never blended with speed. Pick error rate running above 1.5× the warehouse average for two consecutive weeks triggers the improvement process. Errors cost 3–5× what slowness costs — one mispick, and the return, reship, and customer-service time eat half a day's wages.

Line three: attendance. Two unexcused absences or four tardies inside 90 days triggers a conversation; do it again and you're out. Attendance is the one metric with zero sentimentality, because it spreads — if the new hire coasts on chronic lateness, the veterans' discipline collapses within a week.

All three lines go into the onboarding paperwork in writing, signed by the new hire. The scorecard updates weekly, reviewed and co-signed by supervisor and employee. This paperwork isn't bureaucracy — it's the only shield you have when you terminate.

The red lines of termination: cutting corners costs the most

Most US states are at-will employment — in theory, you can fire without cause. But "in theory" is the most expensive phrase on a lawyer's invoice.

I watched a Texas warehouse learn this the hard way. A supervisor, wanting it over with, fired an underperforming new hire on day 85 of probation. Problem: the worker had filed a workers' comp claim two weeks earlier. The employee's lawyer needed one sentence: "Fired two weeks after a comp filing — you call that a coincidence?" The settlement ran past $40,000 — enough to hire 20 new people.

Three red lines:

First, the documentation line. Terminating without a written performance record is walking in naked. Discrimination claims flip the burden — the employee only has to show they belong to a protected group and were fired, and then it's on you to prove the firing had nothing to do with it. Your evidence is those 90 days of scorecards, conversation notes, and improvement plans. Every missing page raises the legal bill.

Second, the timing line. The employee just took medical leave, just filed a comp claim, just complained about a supervisor — and gets "failed probation" the next day? No judge or jury believes in that kind of coincidence. When the timing is bad, wait two weeks, build the file properly, or call your lawyer first.

Third, the consistency line. The reason has to be identical everywhere. "Don't come in tomorrow" to the employee, "failed probation" to the unemployment office, "attitude problem" to the lawyer — three versions, and the plaintiff's attorney triples the price.

Probation rules, termination notice, and final-pay requirements vary widely by state — follow your state's current statutes and your counsel's advice. My only financial guidance: budget $50,000 as the expected cost of a single wrongful termination. Not because it will happen, but because one occurrence wipes out the savings of 20 correct hires.

The raise formula at conversion

Conversion isn't "90 days pass, everyone gets a bump." It's a repricing event. My formula:

Conversion raise = throughput factor × quality factor + attendance bonus

• Throughput factor: 1.0 if 90-day UPH ≥ 95% of seasoned average; 0.6 for 80–95%; below 80%, no conversion.
• Quality factor: 1.0 if error rate beats the average; 0.7 if it matches; above average, no conversion.
• Attendance bonus: +1 percentage point for a clean record; nothing with absences on file.

Set the base raise at 3–5% (illustrative; calibrate to your local market). A new hire who nails all three gets 5%; a borderline one gets 2% or converts with no raise. Spend the money where it earns, not as an automatic handout at day 90.

The calculable conclusion: one wrong hire costs roughly 4–6 months of that role's wages, and every 30 days earlier you stop the bleeding saves about one month's wages. So the KPI of probation shouldn't be the conversion rate — it should be the day-30 identification rate. The faster you spot a bad fit, the more the whole system is worth.

By Lao Zhou, Logistics Cost Analyst