Optimizing Inventory Management
Some industries are really good at certain things. Maybe you [CPAs] can talk to client business executives and say, “You know, in the Grocery industry they have 20,000 SKUs and a profit margin of only 3%, so they better understand how to move product, a lot of which spoils, and be really on the money in terms of inventory management.”
If you work with a small business with, say, 12 SKUs, consider inventory management the way a grocer must. Take some of the sophistication that is used for a very large quantity of SKUs, which no one can manage in his or her head, and apply it at your organization where someone might simply decide, ”It looks like we could use some more X, let's get some more." Now, that reordering point—where did that number come from?
In inventory optimization, we are taught that there are generally three numbers emphasized: What is the demand? What is the variability of the demand? What is the replenishment time? To make managing inventory easier, you might calculate a minimum level and a reorder level. You might say that when a SKU falls below 15, reorder 10.
Your question should be: When was the last time those numbers were calculated? An executive may respond, “I set 15 as the minimum level because that's how my Dad and my Granddad did it. Then I upped the level because we ran out once.” Or perhaps the person who set those levels was a professor who analyzed the inventory turns and provided those numbers. In any case, maybe it is time to revisit that basic formula.
Optimizing Resources at Service Organizations
Most service organizations would say their labor costs are basically their raw materials, but they are not uniform. For example, an organization with different classes of labor may need one experienced supervisor for three employees on the day shift, but at night there are different ratios. Figuring out all the rules to handle peaks and to minimize labor costs tends to be a good optimization problem.
We have found that a lot of organizations have rules of thumb to staff but they have never really questioned them. Executives will say something like, “You need one experienced manager and one assistant manager and three servers,” but they've never really questioned this rule. I believe you can use analytic models to simulate peaks.
One Big Data app that many people are thinking about in all sizes of service organizations is, “How much business are we losing when we can't handle it?” At the micro level, you can easily look at sales you made, but it is difficult to see when a customer walked in and there was no salesman available, so the customer had to walk out. At professional organizations such as accounting or consulting firms, many of us are presented business development and sales opportunities that often are being chased down by the senior staff and partners, and it would be state of the art to dispassionately employ a scoring system-- before the lead comes in—to assess how likely a piece of business will be won.
This post is adapted from a presentation and discussion with the New Jersey Society of CPAs (NJCPA) in October 2014.