3 ms·
In a multi-stage process, even if you fix the cause with 80% responsibility, the outcome may not improve by much, as there is failure now at another stage. Of c
by enugu 2y ago
In a multi-stage process, even if you fix the cause with 80% responsibility, the outcome may not improve by much, as there is failure now at another stage. Of course, fixing that single cause is still progress.
- vasco 2y agoThen it didn't have 80% responsibility and you failed to identify the bottleneck, it just means you identified a local optimisation, which is again "every little bit helps" thinking - what you should do when thinking about queueing systems (or multi-stage) like you describe is to identify the bottleneck. An improvement on the bottleneck should directly translate to an improvement to the end result.
- enugu 2y agoThe issue is when there is overdetermination - fixing a single cause resolves 80% of the current cause of failures, but the process fails at another point which becomes the new bottleneck. Say, most people are failing at a task due to stress(A). But even if you solve that, then most people can start failing due to another cause, B, say laziness. Fixing A is not "a little bit" and any solution will have to face A(in order to have success rate > 20%) but it is not sufficient to fix stress alone. You can of course agglomerate A & B into a single cause C but for that to be meaningful C has to be the 'root' of A and B, otherwise the solution for C also just becomes an agglomeration of the solutions for A and B.
- vasco 2y agoAll I said is you identify the major factor first and fix it, and keep doing that. What you're saying now is basically what I started with, except applying a recursive step of "keep identifying new bottlenecks". But the algorithm should always be "Look at the biggest factor and nothing else", of course you should check if the biggest factor stays the same over time