EXPERIMENT 02 / WORK MEASUREMENT
How long should
good work take?
A stopwatch gives you a number. A good study helps you understand it. Build a labour standard, question the assumptions, and see what changes.
CHOOSE YOUR STARTING POINT
Synthetic examples, not industry benchmarks.
Your changes stay in this tab. Export to keep a copy.
01 / OBSERVE THE WORK
Start with the whole story.
Every cycle is a little different.
Select a bar to investigate that cycle. Keep unusual observations until you understand what happened.
What does the uncertainty tell me?
The 95% t interval estimates the mean of independent, representative cycles from a stable process. Small, skewed or dependent samples can make it misleading. It excludes uncertainty in rating, allowances and the proposed method. It is not a range for 95% of individual cycles, or proof that a standard is fair.
LOOK A LITTLE CLOSER
New rows copy this cycle as a starting point. Replace the times with your observation. Maximum 30 cycles.
How I would run a time-and-motion study
- Agree the task with the people doing it. Define the unit, start and finish, method, equipment, and quality requirements. Observe with their knowledge and participation.
- Break work into observable elements. Separate travel, handling, processing, checking and delays. Record machine time separately; a pace rating belongs only on worker-controlled work.
- Sample the operation you actually need to represent. Cover different trained operators, days, shifts and demand conditions. Keep different task mixes or methods in separate groups.
- Record context alongside time. Note interruptions, rework and unusual conditions. Study motion, layout and safety before asking people to move faster.
- Analyse variation, then validate. Review the distribution and sequence, justify exclusions, calibrate ratings and define allowance conventions. Trial the proposed method with operators and check quality and fatigue.
- Maintain the standard. Record its owner, version, scope and review date. Revisit it when layout, equipment, task mix or quality requirements change.
02 / RETHINK THE METHOD
Change the walk.
Keep the care.
What if supplies were closer? Try reducing travel while keeping the hands-on work and quality check intact.
Press play to follow one task. You can pause at any time.
A what-if assumption. The original observations stay intact. Time the revised method before adopting any predicted saving.
03 / BUILD A LABOUR STANDARD
Make the assumptions visible.
From seconds to a working day.
100% = reference pace. Above 100% means the observed person worked faster than that reference, so normal time increases. Requires trained, calibrated judgement; this example applies one rating to all manual elements. Recorded delays are not pace-rated.
Personal needs, fatigue and unavoidable delays need an agreed basis. 15% here is illustrative. Do not include the same recorded delay again in this allowance. The two conventions give different results for the same percentage.
Fixed work is a total team allowance-inclusive budget, such as opening and closing. Available hours exclude unpaid time. Do not subtract again the personal, fatigue or delay time already covered by the cycle allowance.
YOUR WORKING ESTIMATE
A daily workload equivalent. A usable roster still needs interval demand, skills, breaks, service levels and minimum coverage.
04 / FOLLOW THE MONEY
A few seconds.
A year of consequences.
What happens to the annual labour budget when a study overstates or understates the time the work needs?
Put a price on the assumptions.
All amounts are AUD. These starting values are illustrative, including the wage; enter your own average rates. Team and pay settings stay when you reset or switch the study example.
Allow here for paid leave, absence or work outside this model. Do not subtract breaks or delays already covered by the time-study allowance. Available hours include that allowance.
Explore the overtime assumption
This is a what-if rate, not an award entitlement. It multiplies the base wage and the same on-cost factor. Actual overtime rules and applicable on-costs may differ.
ANNUAL BUDGET OVERSTATEMENT
Three numbers. Different questions.
A budget variance is not an automatic payroll saving. Headcount and paid hours drive the existing team’s annual pay. A flawed standard can change planned staffing or hide a shortfall; spending changes when hours, hiring, overtime or deployment actually change. Costs of service failures, rework and turnover are not estimated here.
See the annual cost calculation and boundaries
Both standard comparisons use the current daily volume and fixed-work minutes from the study above, repeated across the operating days. Only variable task time receives the error percentage. Annual staffing equivalents allow fractional employees; hiring in whole people, seasonal demand, task mix and roster constraints need a separate plan.
Employer on-costs are a single editable factor. They can include superannuation, workers compensation and applicable payroll tax; this example does not calculate those obligations individually. Paid leave already in paid weeks should not be added again as extra wages. See the Australian Government’s hiring guide ↗ for the cost categories to consider.
No pinned or synthetic standard is automatically correct. The comparison shows sensitivity to assumptions and does not prove that a study is good or bad.
05 / MAKE THE CALL
A neat number.
But would you use it?
Try a judgement call, then build a practical review checklist for your own study.
THE OBSERVER’S DILEMMA
One cycle takes much longer. What next?
Choose an approach to explore the trade-off.
Evidence I would want to see.
0 of 6 review prompts considered.
TAKE THE QUESTION WITH YOU
What would change your confidence in this standard?
Export your assumptions and observations, or copy a discussion starter for a conversation with your team.
Can the system use the study?
A study measured by departments X, Y and Z may need a different activity structure for forecasting. Explore the mapping gaps and the potential cost of model rework in the implementation-gap experiment →.
Connect each validated task standard to a volume driver: orders, cartons or tables. Separate fixed work from variable work, account for task mix, and distribute workload into planning intervals. In a workforce management implementation, agree those definitions and their ownership before mapping them into the configured labour model. Product capabilities and configuration vary; this playground is a conceptual model.
Compare notes with Jasper →