Work Sampling for Headcount Justification
Use observations to analyse staffing needs alongside demand, quality, safety and service variability.
Updated:
Activity distribution can inform workload and constraints. It does not alone determine optimal headcount or justify job reductions. Combine it with demand, time requirements, quality, skills, safety, absence coverage and break rights.
What is measured
Working-time records show daily start and end; sampling estimates states within the observed scope. It does not reconstruct precisely what every person did throughout eight hours. A participation is not a measured minute.
Define direct work, necessary support, confirmed waiting, breaks and unobservable states. If direct attention is 43.2%, the other 56.8% is not automatically spare capacity: it may include support, training or system-imposed waiting.
Campaign design
Define population, positions, period and exposure. Pilot categories and cover relevant conditions. Explain the study and document familiarisation, omissions and delays. Do not camouflage observation or claim unpredictable timing prevents all reactions.
For an independent simple proportion, n₀=ceiling(z²p₀(1−p₀)/e²). At two-sided 95% and p₀=0.5, this gives 385 for five points and 1,068 for three. At 99%, z≈2.576 and five points, it gives 664. These are planning assumptions, not independence or coverage guarantees. Wilson and advanced calculations still require design assessment.
From activity to capacity
Estimate category hours using weighted share×compatible person-hours. Do not add absences already excluded or assume simultaneous waits each create separately recoverable output. Staffing also needs demanded workload, task standards, availability, coverage and service constraints.
High waiting shares can prompt logistics investigation; they do not prove hiring is unnecessary. Bottlenecks, relief cover and simultaneous demands may require another solution.
Simulated example
A fictional 85-person department has 51% direct attention, 15% quality waiting and 9% tool searches. After assumed changes, those two waits fall to 5% and 3%. The sixteen-point difference might be partly reassigned, with direct attention assumed at 67%.
These are teaching figures, not Worksamp results or proof that 20% more orders can be absorbed without hiring. That requires output, demand, load and costs. Removing quality controls for speed is not automatically acceptable improvement.
Labour and technical limits
There is no general legal duty to use work sampling or automatically superior evidentiary value in restructuring proceedings. Relevance and sufficiency depend on the actual procedure. Sampling does not replace ergonomic/psychosocial assessment or individual working-time records.
It can support availability, not every OEE factor without counts and ideal cycle. Nor does it guarantee anonymity: assess actual data processing.
Worksamp supports activity-study planning and capture. Use results as one part of competent analysis and appropriate participation, not a verdict on individual performance.
Sources: NIST, Spanish Workers' Statute. Worksamp.