This post could be a sequel of “Yeah, problem solving” in which I used Peter Senge’s quote: “Today’s problems come from yesterday’s solutions”.
Quite often people we consultants meet are puzzled by a problem they can’t understand:
- a reliable process or machine suddenly seems out of control,
- steady performance dropped unexpectedly and with no apparent reason,
- sudden quality issues with trusted supplies,
Our experience lead us to investigate the last change made, precisely because of the “wisdom” of Peter Senge’s quote: chances are that a modification (fixing a problem) led to unexpected Undesirable Effects and causing new a problem to appear.
Of course, the modification to look for is seldom the worried person’s ones, which he/she would most probably remember and perceive the possible cause-to-effect relationship.
No, the modification more likely happened outside the span of control and without the knowledge of the impacted people.
A modification leading to a problem in a lengthy process can happen far away (both in process steps and location) from the point the problem appears, letting the people perplexed about this reliable process now out of control.
Purchasing and procurement choices are unfortunately often the unintentional culprits, buying a slightly different grade of material, changing a supplier or accepting a low quality batch with the best intentions: cut costs or ensure timely deliveries.
When facing a puzzling problem the investigation should follow “the last modification path”.
This isn’t always easy though. The Undesirable Effects brought up by the change may be minimized or even neutralized for a while, long enough for everybody to forget about the nature of the change, when it happened and its consequences then.
That’s precisely why some industries with strong safety and regulatory constraints like aeronautics or pharmaceutical have to be cautious about any modification (needs approval after thorough risk assessment) and capture every information about virtually anything (dates, manufacturing conditions, persons in charge, certificates…), in case an investigation must find the root causes of a deviation (or worse), long time after the triggering action occurred.
When the problem cannot longer be neutralized by the former forgotten fix, it looks like a new problem.
Searching for the last change is often a good guess, yet not always leading to the root cause. Keep in mind that some modification correlate nicely with the apparition of the problem, but correlation isn’t causation.Follow @HOHMANN_Chris