Process Analysis
Process Analysis
Process analysis is where you prove, with numbers, where a process underperforms and why. The exam tests your grip on performance measures, value-added and waste analysis, root-cause tooling, and the arithmetic that appears in scenarios - Little's Law, value-added ratio, FMEA risk priority numbers.
Measure the process, not the activity
- Cycle time - elapsed time from start to finish for one instance, including waiting and rework. Processing (touch) time is only the working portion. Lead time is often used for the customer-visible span.
- Throughput - instances completed per unit time. Capacity - maximum sustainable throughput. Utilization - throughput divided by capacity.
- Little's Law - WIP = throughput × cycle time. Rearranged, it answers how much cycle time falls if WIP is cut.
- Quality - first-pass yield, defect rate, DPMO, rework hours.
- Value-added ratio - value-added time divided by total cycle time.
Waste and root cause
The eight wastes (TIMWOODS): transport, inventory, motion, waiting, overproduction, overprocessing, defects, unused skills. Root-cause techniques: 5 Whys, fishbone (people, machine, method, material, measurement, environment), Pareto to stratify, FMEA with RPN = severity × occurrence × detection, plus process mining and discrete-event simulation where event data exist.
Worked example
An order process takes 10 minutes of processing but 8 hours (480 minutes) of elapsed time. Value-added ratio = 10 ÷ 490 ≈ 2%. A team proposes cutting processing time by 20%, saving 2 minutes. The analysis says hold: waiting dominates, so removing a queue or approval moves cycle time far more. Separately, 200 orders sit in progress and 50 complete per day, so Little's Law gives cycle time = 200 ÷ 50 = 4 days - a figure you verify against timestamps rather than trusting an estimate.
Traps that cost marks
- Reporting averages as if they describe every case; variation and percentiles drive customer experience.
- Confusing processing time with cycle time, then claiming improvement that never touched the wait.
- Jumping from symptom (late orders) to solution (more staff) without root cause.
- Misordering FMEA factors or scoring detection as if higher is better - higher detection difficulty is worse.
- Ignoring rework loops, which add time and cost invisibly.