Email
sinha.rk.dr@gmail.com
How should we treat ‘outliers’ in Supply Chain?
Supply Chain teams often struggle with outliers in their data. These are found in all data sets… demand data, supply data, production data. The most common treatment meted out to these outliers is to ignore them as if they didn’t exist. We ‘clean’ the data as if these outliers were undesirable elements which deserved to be purged out of the system. Is it the right thing to do?
I have often wondered why we call such data points as outliers. Is it because they don’t match with our assumptions? Is it that our current logic is incapable of explaining these data points? Are we refusing to see and accept reality as it is?
In my opinion, we should first stop calling them ‘outliers’. These are actually gold mines which signal to us that our current assumptions are either wrong or incomprehensive to explain them. If we develop curiosity to look deeper and try to understand them, our logic would expand and sharpen.
Next time we encounter such an ‘outlier’, we should give it due respect, acknowledge its presence and question our current assumptions and logic.
Remember that some of the greatest scientific discoveries, like presence of the elementary particle ‘Higgs Boson’, are based on observing such outliers.