Tuesday, March 24, 2009

Alternate Perspectives on Supply Chain Costs or .. What Do You Get When You Cross Eli Whitney with Frankenstein?”-Michael Sachs & Brian Crounse


Most people know of Eli Whitney as the man who invented the cotton gin – the big invention of the industrial revolution. Another, more important legacy of Eli Whitney was the popularization of interchangeable parts (something we all know and love about the service-parts business).

Times like these have us all thinking about how to save money, particularly within our supply chains. This week, we’re headed for the lab where we’re going to use Eli Whitney’s interchangeable parts to build a couple of supply chain Frankenstein monsters. We will do this by swapping like-kind parts, but they are going to come from different OEMs. So, what we end up with is a supply chain that takes the best (or average) of the OEMs and pieces them together into one supply chain monster.

To keep things simple, let’s use four fairly large parts of the service-parts supply chain: 1) outbound transportation, 2) fixed facility costs, 3) variable warehouse costs, and 4) inventory carrying costs. We’re going to ignore inbound transportation due to the huge impact imports have on their cost differences. So think of our monsters in terms of the flow of goods to the customer once initially landed at one of our warehouses.

Let’s start by building our Heavy Equipment monster – our “HE Frankenstein.” The range of overall supply chain costs (expressed as cost as a % of sales) across the industry exceeds a 2-to-1 ratio, high-to-low (even after lopping off the lowest cost OEM). Some of this variability is due to company-specific opportunities and constraints, but a lot of the variation is driven by performance. So, what if we take some of the better parts from various OEMs and build an HE Frankenstein?

For each part (outbound transport, fixed and variable warehousing, and inventory carrying), we’re going to compare cost as a % of sales. This neutralizes the impact of volume on costs. We’ll then look throughout the industry segment for the part we’d like to use for building HE Frankenstein. Once we’ve selected a part, we’ll figure out the cost of sales differential (relative to the rest of the industry) and apply that number to actual costs. This will enable us to monetize the savings for the industry.

Outbound transportation is one of the highest cost elements of our monster. Looking once again at costs as a % of sales, we find that for outbound, the high cost OEM is nearly 3 times higher than the (comparable) low cost OEM.

Outbound transportation is an area where there is a lot of commonality. All of the OEMs have dealers and distributors across the country. All of the OEMs generally offer similar service levels, with stock orders delivered in 2-5 days and emergency orders delivered within 24 hours. The big differences between companies lie in network density (dealers and depots) and shipment policies (e.g., referral chains, will-call availability, etc.). There is also a mixture of two-tier and single tier networks, with most OEMs somewhere in the middle. What’s interesting is that these differences in structure don’t explain much of the difference in costs. In fact, significant contributors to the difference include mode choices and freight rates. Freight rate benchmarking has often demonstrated that rates are very inconsistent, even when viewed by mode on a specific origin-destination pairing.

If everyone were to perform at the best-in-class (BIC) level for outbound transportation, OEMs could save a total of more than $115 million annually (or an average of nearly $15 million for each non-BIC company). But that’s pretty ambitious. Let’s also look at moving to median-level performance all companies that have higher-than-median costs. The result is a savings of nearly $60 million annually across the industry (or an average of $12 million for each company above the median line).

Fixed facility costs are mostly a function of how many facilities we have, where they’re located, whether they’re owned or leased, and, if owned, how long they’ve been depreciating.

Variable Warehouse costs are driven by two primary components – labor cost per hour and productivity. For this analysis, we correct for labor cost per hour, by giving everyone the median wage of the industry. This way, wage differences are neutralized and will not interfere with the cost savings analysis.

The following table summarizes the effect of interchanging four good parts for the four not-so-good ones. You’ll also see that we don’t even need the best parts to save hundreds of millions of dollars. In fact, we just need to use average (median, really) parts. *Best-In-Class is based on using the data for the 2nd or 3rd-best performer; some top performers operate under unique circumstances.


So, if everyone were to use best-in-class parts, the industry could save $375 million. While that may seem unrealistic, one would be hard pressed to argue against a goal of mid-pack performance. But by using mid-pack parts to build HE Frankenstein, the industry could save well over $200 million.

Before talking about some of the things you might do to achieve mid-pack performance, let’s do the same rebuilding of the auto monster – “Auto-Frankenstein.” As with HE, we’ve lopped off the lowest cost provider before quantifying best-in-class.
*Best-In-Class is based on using the data for the 2nd or 3rd-best performer

If Auto Frankenstein were to perform at the best-in-class level, OEMs could save a total of $1.3 billion annually. Drawing a line at the median and calculating the potential savings for the OEMs above the line results in $535 million in annual savings. That’s over a half-billion dollars just to get to mid-pack!

Now before you email or call us with your concerns regarding lack of part interchangeability or the interconnectedness of parts (e.g. adding warehouses increases fixed and variable costs while reducing transportation costs), I will admit that we have not built the perfect supply chain monsters. Besides, we’re talking about Frankenstein – not exactly the most elegant (or manageable) specimen. But the fact remains that no one OEM is best at everything. Even after correcting for the reasons why any one company is “different” we’re going to find that the opportunities are huge.

So, what can OEMs do to build their own Frankenstein monsters? There are two approaches here. One approach assumes that you can’t re-build all of the supply chain parts yourself. In these cases, you would be better off working with other OEMs to build the part that is interchangeable with both organizations. This is happening today. Some OEMs are collaborating with other OEMs on outbound transportation. By combining routes, route densities increase and, therefore, delivery costs go down. With sales going down, inventory will be going down proportionally. This means warehouses will have empty space. Are their collaboration opportunities that would enable better utilization of fixed facilities?

The other approach is to build your own monster parts. Audit freight bills not just for accuracy, but also to find savings (they’re in there). Benchmark your freight rates – based on our recent analysis, there is a surprising lack of correlation between volume and rates on a number of high volume lanes. Rethink some mode choices (and associated order response times). Increase warehouse productivity by up to 50% (for those of you that attend our conferences, you’re aware of at least two OEMs that have done this). Stop ordering so many parts with high likelihood of obsolescence. And for those of you that collect dealer inventories (and if you don’t, you should), why do you continue to order additional inventory on some parts when you have more than a lifetime supply in the field?

In these economic times, we can not afford to play it safe. Ferret out every practice in your supply chain that is done simply because “it is the way we have always done it.” It can be hard to change, to admit that others may have a better idea. Today, however, it is the way to survive.

"Man," I cried, "how ignorant art thou in thy pride of wisdom!" – Dr. Frankenstein

Please let us know your thoughts on our blog CarlisleBlogFeedback.com

No comments: