Moneyball Logistics

I read the book “Moneyball” by Michael Lewis and loved it. It was about how the Oakland Athletics, a baseball team with a small payroll, bucked conventional thinking and made themselves a consistent playoff team in the late 1990s and early 2000s while competing against much wealthier teams with much larger payrolls. Their ability to win was based on a form of statistical analysis invented by Bill James.

In short, this book put math in a leading role. But because Michael Lewis is a terrific writer, it was still an easy book to read. I was also amazed when Brad Pitt made this book into a movie. I would have bet money that you couldn’t make an interesting movie out of this book, but I would have been wrong.

I recently heard Andy Andres, a Senior Lecturer at Boston University who studies this form of math, speak on this topic at the ToolsGroup North American User Forum. Andy said that if you fast forward fifteen years from when the A’s first began using these techniques, and Bill James was considered a bearded freak, every team in baseball except for two have put this form of analysis at the center of their front office operations.

He also provided some extra details on the math behind Moneyball.

Stay with me here. I’m going to provide a short overview of Moneyball math, and then link this form of analysis to logistics.

Here’s the key formula driving front office decisions today: A team’s winning percentage will be very close to Runs Scored squared (RS) divided by the sum of Runs Scored squared plus Runs Allowed (RA) squared. The time period for RS, RA, and the winning percentage is a season.

Andy ended his speech by admitting he knew nothing about supply chain management. But he challenged us to come up with some form of this formula that would lead to winning supply chain organizations.

I can’t claim I know the formula. And I think the formula will be somewhat different for every company. But I think I know how to get there.

Here is how I would go about creating this formula. I would start by conducting a study to quantify the cost of customer dissatisfaction resulting from order errors. Specifically, I would want to know what deviations from the perfect order are costing the company in terms of lost customers and the customer’s share of wallet. For a description of how to do this, see the article titled “Competing on Supply Chain Analytics.” What makes this more complex than the Staples story in the article is that for many companies optimal customer responsiveness levels will be different for different customer segments and sales channels.

Knowing the cost of poor quality associated with perfect order metric failures tells a company how much money it can realistically spend improving its order fulfillment capabilities. The handful of companies I know that have conducted various types of “cost of quality” studies  have found that errors generally cost them far more than they thought it would.

So now you have two parameters of the formula, service level and supply chain costs.

Initially, you can increase service levels while decreasing supply chain costs by implementing supply chain applications. For example, companies that implement an inventory optimization solution will find that the solution allows them to improve service levels while holding less inventory. Warehouse management systems improve order fulfillment accuracy while reducing warehouse labor costs. Transportation management systems can help get orders to the customer on time while reducing freight costs.

But companies will ultimately get to a point where improved order fulfillment accuracy will involve higher supply chain costs. But now they will have a formula to let them know how the service level/supply chain cost tradeoffs will impact revenues and profits.

Just like with Moneyball math, early adopters will gain competitive differentiation. In the longer run, however, this math will become known and almost all companies will understand and make use of it.