As long time SCDigest readers will know, we have off and on done a number of pieces on "probability" thinking and simulation, especially with regard to estimating project ROI. More again on that soon.

So, we were very interested to see a recent article from the McKinsey Quarterly, which describe how a large global auto manufacturer used probability analysis to better assess supply chain risk.

As shown in the graphic below, the company used Monte Carlo analysis to simulate both demand and supply capabilities. This was in contrast to the "single forecast" type of approach the company had used in the past, even when looking at that demand under different discrete scenarios.

This approach led the group to generate a probability distribution of demand (by geography and by product) that together included some 15,000 scenarios. To this bell curve, the team mapped the ability of the company’s production network to meet potential demand profitably in each scenario.

Source: McKinsey Quarterly

As the chart shows, using this analysis, the company could see that the probability was that it would forego significant revenue by not having the capability meet the most likely demand scenarios.

"Certainly, much of this amount was impossible to capture and always would be—only one demand scenario would prove correct, after all, and production resources are finite," McKinsey observes. "Nonetheless, armed with this information, the executives could now begin to look for ways to increase the company’s operational flexibility on the margins to capture more of the upside if demand proved higher than expected."

Have a comment? Send it at the Feedback button below.