Common Mistakes

Why Your Labor Standards and Crew Numbers Are Wrong: Common Mistakes

A troubleshooting guide to the most costly errors in labor standards and skills planning, from allowance mistakes and unit mixups to stale time studies and phantom skills coverage.

Symptom: your standard minutes per unit look tight on paper but the floor always runs over. Root cause is almost always a missing or wrong allowance. Engineers time the observed cycle, get 4.2 minutes, and publish it, forgetting the PF and D allowances for personal time, fatigue, and unavoidable delay. A typical allowance runs 12 to 18 percent, so a 4.2 minute observed time should become roughly 4.9 to 5.0 standard minutes. Skip it and every crew looks 15 percent short. Fix: never publish a raw observed time from the Standard Minutes Per Unit calculator without layering the allowance in, and document the percentage you used.

Symptom: two shifts running identical work post wildly different productivity. Root cause is usually an unnormalized performance rating. During a time study the analyst must rate operator pace against a 100 percent normal, so a fast operator observed at 3.6 minutes might rate at 115 percent, giving a normal time of 4.14 minutes. Forget to rate, and you bake one person's pace into the standard for everyone. Fix: rate every study, keep ratings within 85 to 125 percent, and throw out any observation where you cannot defend the rating. Feed normalized times into the Labor Standard Calculator, not stopwatch raw data.

Symptom: crew size math says you need 6 people but the line chokes at a single station. Root cause is averaging cycle times instead of balancing to the bottleneck. If five stations run 40, 45, 38, 62, and 41 seconds, the line paces at 62 seconds, not the 45.2 second average. Sizing crew off the average understaffs the constraint and strands 27 percent of takt. Fix: when you use the Crew Size Calculator, drive headcount off the slowest station and takt time, then rebalance work content so no station exceeds takt by more than 5 percent.

Symptom: direct labor cost per unit swings 20 percent month to month with no process change. Root cause is a wage rate that omits the fully burdened cost. Teams plug in the 22 dollar base rate and forget payroll tax, benefits, and paid time off, which add 30 to 45 percent. A 22 dollar base is really 29 to 32 dollars loaded. Fix: always feed the fully burdened rate into the Direct Labor Cost calculator, and reconcile it against actual payroll dollars divided by clocked hours each quarter so the loaded rate never drifts.

Symptom: your indirect labor ratio looks healthy at 0.25 but overhead keeps climbing. Root cause is misclassification, where working leads, material handlers, and inspectors get coded as direct when they are indirect, or the reverse. Moving 3 handlers out of a 40 person direct pool changes the Indirect Labor Ratio denominator and can shift the ratio from 0.25 to 0.32. Fix: audit the labor coding in your timekeeping system against actual job duties twice a year, and hold indirect below 0.35 of direct for most discrete manufacturing before you trust the trend.

Symptom: the skills matrix shows full coverage yet one absence halts a cell. Root cause is counting anyone with a 1 as qualified when your scale conflates awareness with proficiency. If coverage on a critical task reads 100 percent but three of four operators sit at level 1 of 4, real independent coverage is 25 percent. Fix: in Skills Matrix Coverage and Cross-Training Coverage, only count level 3 or higher as coverage, and require at least 2 fully proficient operators per critical task so a single vacation does not stop the flow.

Symptom: training plans blow their hours and new hires still miss standard at week 8. Root cause is estimating training hours off classroom time only and ignoring the learning curve to standard. If a task carries a 90 percent learning curve, the operator needs many repetitions before hitting the standard minute value, and unit two takes 90 percent of unit one, not full speed. Fix: build Training Hours Required from time to proficiency, not seat time, and expect a new hire to reach standard around the 200th to 400th cycle on a moderately complex task.

Symptom: onboarding cost looks cheap so you churn people freely, but the P and L disagrees. Root cause is counting only recruiter fees and the first-day paperwork while ignoring the productivity ramp and trainer drag. A new hire producing at 60 percent for the first 6 weeks, plus a trainer pulled to 70 percent output, hides thousands in lost throughput per head. Fix: load the ramp loss and trainer time into the Onboarding Cost calculator, where a fully loaded figure of 4,000 to 8,000 dollars per operator is common, then use that number to justify retention spending.

Published 2026-07-01.