Well during a recent kaizen event I discovered a most interesting corollary to this theorem. Although it doesn’t seem to matter whom you ask, it does matter who’s asking. The +50% rule appears to apply only when an engineer is asking the question. If a manager asks the same person the same question, they get a totally different number. Early data suggests that it is actually 50% less than actual. I was fascinated to find this out, and I wish I could pretend to understand the psychology of why this is.
Nevertheless this has profound implications for gemba management. If this theory holds (and clearly more study is needed), then we might be able to completely eliminate the need for time observations for standard work. All we have to do is have an engineer ask an operator how long it takes them to complete a task. Then have a manager ask the same person the same question. The actual number will be exactly halfway between the 2 answers! Guaranteed! I think there’s a paper in this.