“Lies, Damned Lies, and Statistics” — WTF?
I guess you could say the perceived use and value of statistics has changed a bit since Mark Twain popularized this phrase in the early 1900’s. Today, there’s no disputing the power and value of statistics in business, though it is woefully underutilized by small and midsize businesses. The redemption of statistics from Mr. Twain’s satirical antipathy began in earnest during WWII. Here is one of those stories that Americans can be thankful for.
Statistics in the Trenches
Around 1933, the brilliant statistician Abraham Wald immigrated to the US and was recruited to help with the US war effort by the Statistical Research Group. Early on, the group was presented with photos of returning battle-tested planes riddled with bullet holes and asked to find a design solution for armoring future planes. Most holes were in fuselages, wings and tails with few in engines. The military wanted more armor in the areas with the most bullet holes and asked how to better protect planes. Wald famously gave a counter-intuitive answer to their query. He explained that you don’t want to put armor where the bullet holes are. The armor needs to go where the bullet holes aren’t, which was on the engines. Wald insight, which today is called survivor bias, save many lives by focusing on why planes were not returning, which was due to bullet holes in engines.
What Does This Have to do With Business?
There are many lessons in this story. One is that there is frequently more value in dissecting failure than success. When analyzing business failures, such as customer losses or employee departures, take a “lessons learned” approach and then implement remedial solutions. Another is the importance of statistics to continuous improvement. Like in war, in business small improvements have out-sized effects.
Customer Satisfaction and Statistics are Highly Correlated
An example in business relates to product service levels and reorder points. A service level for a given product is the probability that it is available when ordered. A 95% service level also implies a 5% chance of stockout. Applying statistics to historical demand and lead time, one can compute a reorder point that ensures a 95% service level. Utilizing this statistical technique with clients, we bring greater confidence to their replenishment efforts and greater control to inventory levels, helping them make better decisions that drive profitability and cash flow. No lie! Contact us today to learn more.
If you are a business owner that is looking for ways to leverage your data analytics, please contact us today.