14. Interview with Mr. Mike Orr of Genuine Parts Company
Introduction
This chapter consists of a series of questions and answers from an interview with a leading supply chain executive. The objective of this chapter is to provide a practitioner’s perspective on the subject of life cycle issues in supply chain management. Other related subjects are also discussed in the interview.
About the Interviewee
Mr. Mike Orr is Senior Vice President of Operations & Logistics for Genuine Parts Company. Mike joined the company in 2005, and has more than 25 years of experience in global operations for industrial and consumer products companies. During his time with Genuine Parts Company, he has also served as Senior Vice President of Operations & Logistics for S.P. Richards Company, a GPC subsidiary.
Prior to joining Genuine Parts Company, he was Group Vice president of Operations for Newell Rubbermaid, where he had global operational responsibility, including manufacturing, safety, supply chain, and logistics. Before joining Newell Rubbermaid, he worked for Allied-Signal (Honeywell), in a variety of engineering and operation assignments.
Mike earned his Bachelor’s degree in Mechanical Engineering from South Dakota State University and his Master’s degree in Business Administration from Arizona State University.
“Automotive parts is a 30-minute business. If you can’t deliver the part in 30 minutes, you won’t get the order.”
—Mike Orr, 2011
About Genuine Parts Company
Founded in 1928, Genuine Parts Company (GPC) is a service organization engaged in the distribution of automotive replacement parts, industrial replacement parts, office products, and electrical/electronic materials. Their products and services are offered through a network of more than 2,000 operations, geographically located across the United States, Canada, and Mexico. The regionally located distribution centers provide a unique ability to adapt product and service lines to better suit customer needs. GPC’s commitment and reputation for just-in-time (JIT) service position us as a critical partner in our customers’ success. GPC began to diversify its product lines over 30 years ago into several end markets with strong growth opportunities. The Automotive Parts Group, the largest Division of GPC, distributes over 380,000 products through its 58 NAPA distribution centers located across the United States. The Automotive Parts Group also consists of NAPA Canada/UAP Inc., Canada’s leading automotive parts distributor, as well as Auto Todo, one of Mexico’s largest automotive aftermarket organizations. In addition, we operate our heavy vehicle parts distribution business under the name TW Distribution and Traction, and our import parts distribution operations under the name Altrom. The Industrial Parts Group, which operates under the name Motion Industries, distributes over three million items from 500 operations located throughout the Unites States and Canada. Motion’s rapid-delivery model provides customized product and technical expertise to customers in a JIT response, enabling clients to reduce production costs and free working capital. The Office Products group distributes over 50,000 business products from 45 distribution centers across the United States and Canada under the name S.P. Richards Company to thousands of office product resellers, including independent dealers, large contract stationers, national office supply superstores, mail-order distributors, Internet resellers, and college bookstores. The Electrical/Electronic Materials group is one of North America’s leading distributors of process materials, production supplies, industrial MRO (maintenance, repair, operations), and value-added fabricated parts. Primary markets are the electrical OEM (original equipment manufacturers), apparatus repair, and assembly markets. The group distributes more than 100,000 items from 36 locations in the United States, Puerto Rico, Dominican Republic, Mexico, and Canada.
Interview Questions and Answers
How would you characterize your organization’s supply chain?
GPC has complex supply chains across four distinct industries: automotive, industrial supplies, office supplies, and electrical supplies. The product offering and item count is extremely large: 32,000 office products, 400,000 automotive, 4 million industrial, and 25,000 electrical. Our supply chain spans from the source, a manufacturing facility, to the actual consumer of the product. GPC operates some of our own manufacturing sites, and we utilize contract manufacturing partners for selected categories. Of course, we also distribute national branded products that are sourced from parallel supply chains.
We are primarily a North American distribution company, but we ship goods all over the world. Our supply chain source is global. We do have production facilities outside of the United States. Although we ship goods all over the world, in our model we try to keep our inventory close to our customers and their consumption. That closeness is our value proposition. For example, NAPA stores are replenished via our internal supply chain system every night. If you go to a NAPA store today and can’t find a part, we will have it to that stored by 7 a.m. the next day.
How do you deal with the Introduction Stage of a product or product line requiring a supply chain?
It depends on products or parts! Let’s talk about new automotive parts. We class or identify every part or item used to assemble a car. We do this for every car made that is sold or distributed in North America, regardless of where it was manufactured. In essence, we know the future demand for replacement parts for each car that is sold in the United States. We know and keep track of where each part fits on each car, by make, model, and year. For example, a single spark plug “item” can be a replacement for multiple make and model cars and could be used for multiple years. That same spark plug unit may also be used by multiple manufacturers.
We know from our experience that the demand for spark plugs for a new car will be very minimal during the first several years. As we plan inventory deployment, the depth of inventory for that item would be minimized until the projected demand curve grows. In the replacement parts business, the “sweet spot” for demand creation is when a car has been in service six to nine years.
Establishing class is the first step in understanding demand patterns. The second phase is to collect current information for every car registered in the United States and where they are located. This information is readily available, and we categorize location by ZIP Code in the United States. Inventory is then deployed to individual store locations based on probability, age of vehicles, and items that would serve customers in that specific geographic area.
This comprehensive and data-driven information allows us to operate a highly efficient supply chain service for our customers and consumers. For example, we had a request from an individual for a specialized tail light assembly of a car that was more than 40 years old. Our supply chain network had the part, but not in this particular geographic area. We were able to deliver that part within 24 hours, and this is a routine process for our supply chain.
Through our supply chain network, we are able to deliver many types of standard or unique replacement auto parts to customers. Although we are proud of our supply chain performance, this is not a unique supply chain. A similar level of service is provided by companies like Caterpillar and John Deer because they have similar product complexity and need for our responsive customer service.
The challenge for our supply chain is product proliferation. When we get a new product or item in the supply chain, we have to make several assumptions or decisions for the item such as what the probability of product failure is going to be, what the timing (age) of failure will be, based on experience, what is likelihood of a repairs taking place, and finally, what replacement of an OEM or aftermarket substitution is likely to be used.
Historical data highlights the optimal vehicle age interval (six to nine years) for auto repair work in general. We know that the early interval age cars (before six years) tend not to need a repair or are under warranty repair, and late-cycle cars (after nine years) just don’t get repaired. From these data and probabilities, we work up a demand forecast for new products. It is not a perfect science, but the process allows us to maximize our inventory investment across 6,000 locations.
In our business, parts need to be delivered to customers within half an hour or we lose the order. That dynamic drives inventory deployment, store count, and location. Generally, we place our facilities within 3- or 4-mile radius of our customers.
How do you know what to distribute to your customers given a new product?
In the case of a new car, we know some items are consumable and have specific service intervals. For example, a car will quickly need oil filters, a little later need air filters, even later brakes as they wear out. Of course, we factor the probability of collision demand for each part. Given that information we stock the inventory accordingly. We also look at the inventory items in terms of typical consumer good categories of good, better, best.
We have oil filters that are low cost with good performance, a little higher value (that is, better performance), and high end (that is, best performance). These factors are all depend on consumer behavior and can be associated with the kind of car purchased.
As a general observation, people who buy luxury sedans will buy the corresponding high-end products. People who buy pickup trucks or compact cars are more value conscience. Fewer customers will demand the high-end best quality, more the better quality, and considerably more the good quality products. We factor that into our algorithm for forecasting demand requirements to determine the stocking levels. We also use a hub-and-spoke model for stocking in locations. For example, if we have a city that has ten customer outlets for oil filters, we might stock one with the best quality filter, five with the better quality filter, and all ten with the good quality filter. The idea being that if demand for any filter existed in outlets that were not stocked, the item could be quickly transferred to that outlet like a spoke on wheel. This is a technique to have products in the correct geographic area, but not duplicate all items in every store.
How do you do your demand planning for totally new products?
I have worked at organizations where totally new products were being developed, so no frame of reference exists. When you invent something new or attempt to create a new category, you don’t know for certain if or how strong demand will be.
For completely new items, a more rigorous analysis process is required. It starts with sales and operations planning (S&OP) and incorporates a financial pro forma analysis. The business assesses the market potential of the items, investment required to launch and most importantly the return on investment.
Before we even begin the planning process we would have an estimate for the size of the market for the new item. Sometimes you can go to your large retail partners or customers and they can provide some input on sales expectations. However, even with good information and due diligence, you can still end up with a dud of a product, a broken supply chain, and a write-off.
In cases of a less-than-successful product, with proper planning and communication most partners up- and downstream can recoup their investments. Marketing plays a role in successfully introducing new products and must contribute to the S&OP.
Statistical forecasting is useful in planning new products. Existing products or new products that are modeled after existing products can be more accurately forecast using historical data. New and unproven products require careful estimation effort. That is why an S&OP process is required.
We usually take marketing’s estimates for really new products and adjust them down for any potential bias. Mathematically, forecast error is usually reasonable, but bias error can be very large. Marketing or any group trying to introduce a new item will be optimistic. That optimism can translate to bias that creates error in the forecast. Failure to recognize bias errors can prove costly to a supply chain.
For new items, the groupthink forecast can be extremely biased. In my experience, we have had to cut initial forecasts in half, knowing full well that it is better to chase growth in demand if a new product takes off than to deal with overproduction issues. Of course, we take into consideration the production lead times and how fast we could recover in the event of a rapid growth situation. There are many levers to pull if we see that demand is higher than forecast, such as expediting production with overtime. It is more difficult to address product that has no demand or sales potential.
We also protect our supply chain partners with robust and candid communication. They are advised with our initial demand, but we also let them know that forward demand could be erratic. Because uncertainty exists, we plan and discuss recovery actions such as expediting needed production. This helps to reduce some of the risk in the product Introduction Stage.
How do you deal with the Growth Stage of a product or product line requiring a supply chain?
Unplanned growth in demand is always a challenge. In very rapid growth, you have to protect your customers. Sometimes, the need for allocations, even for your very best customers and larger trading partners, is required. There is no real easy way to deal with growth other than chasing demand for a period of time. As an organization, we have a certain amount of surge capacity available to deal with surges in our supply chain. We might be able to use contract suppliers, overtime shifts, and other stretch capacity in our plants to deal with most cases of demand exceeding supply. The success of this depends on the industry you are in, the lead times, and how fast you can react.
For some explosive-growth situations where demand is substantially greater than existing production capacity, the solution is just a matter of working an extra weekend or adding a second or third shift. What is challenging is what I call time-dependent explosive growth. An example is a seasonal product like Easter candy that has to be delivered before the season selling period. In this situation, you have to allocate what you can to your better customers (however better is defined). In a supply chain, we have retailers, wholesalers, and distributors. Basically you have to pick winners and losers based on their value to the firm. In the consumer products industry, Wal-Mart always wins. Generally, what we see is that the retailers are given the product first, then wholesalers, and then distributors. For untested new products, the allocation is the other way around. These are products that you can’t seem to get a distribution channel for because potential partners don’t want to take a risk on a really new product. For products that others don’t want to take a chance on, you can sell them to wholesalers first. They will stock it, and they know that they can sell them to anyone (retailers, other wholesalers, distributors, or consumers).
How do you deal with the Maturity Stage of a product or product line requiring a supply chain?
Generally, our statistical forecasts will catch this kind of situation and adjust the demand level expectations, either higher or lower. It’s best to let the forecasting models bring down the demand expectations. An interesting lesson here is to take the human element out of the forecasts. People have memories and get emotionally attached to certain products. As a result, they can input a natural bias to keep producing a product that is actually leveling off or declining in demand. When the statistics suggest a product’s demand in softening, we have found it best to follow the data. Use the statistical forecasts in these situations without additional human adjustments being included. We try not to intervene with the demand forecasts for “golden zone” types of products that are steady sellers but are somewhat mature and have hit their apex of growth. The demand for these products follows a pattern, so the data speaks.
When your products become mature, do you seek to reduce the costs of their supply chains?
The need to take out costs for a product to make it completive is always present. The issue is how costs are taken out and the impact on the supply chain. Examples of cost reductions include changing from make or manufacturing to an outsource or contract manufacturer model. Firms also work on a raw materials supplier to lower their input costs. Other efforts to reduce costs include investing in automation, improving productivity in production, and reducing costs by reengineering a product and altering the composition of a product (for example, going from virgin plastic to recycled plastic). Cost reduction is a daily focus. However, you need to be aware of the product life cycle. The effort to reduce costs on a product line whose demand is declining might not produce a good return.
How do you deal with the Decline Stage of a product or product line requiring a supply chain?
Every year, we try to identify the products that are obsolete. At some point, you have to kill the product line. We try to be more proactive and place an end-of-life date on these products and communicate that to our supply chain. Sometimes this drives demand to depletion. Setting a fixed date can motivate a sellout, and that is a positive situation for the company.
We also measure GMROI (that is, gross margin return on investment). This is a standard industry measure for inventory productivity. As the inventory turns start to slow, we need a higher level of gross margin to make the product successful; otherwise, we seek to lower the inventory. When the GMROI is consistently below our threshold, it is an indication the product is nearing end-of-life stage. At that point, we sunset the product and put the notice out through our supply chain.
Sometimes, the supply chain is not successful at eliminating our inventory and we are stuck with some exit costs. When this happens, we have to make a choice of writing off the value or sending the product to a liquidation sale. To offset these costs, we establish an obsolescence financial reserve for each product during its life cycle. The method and cost of the reserve will vary based on a company’s position in the supply chain.
A manufacturer will have a lower cost and is more willing to liquidate obsolete products than a distributor is. A manufacturer also has an advantage of not converting raw goods to finished product. Keeping material in the highest echelon or in raw material form is the most economical move for a supply chain. For example, a supply chain has more flexibility to recover cost when they have a piece of steel, because a piece of steel can be converted into a different product from the discontinued product that it was originally to be use in. By keeping materials in raw state, the manufacturer has greater options to reuse it in other products.
Do you have any concerns about the quality of your products coming out of different locations around the world?
In today’s world market, it is a fallacy to question the quality of products, regardless of their country of origin. Today, product-quality expectations are fully vetted and completely qualified before the product is even shipped. There is much information and work put into qualifying suppliers, validating samples, and reviewing production processes. Most suppliers understand quality is fundamental to business. Some of the most sophisticated factories I have seen are in India and China.
A lot of technology has changed in the past ten years or so. What do you envision is the major impact of these technologies, and where will you be heading?
We are continually investing in supply chain system technology. This includes general communications technology. We have a fleet of more than 10,000 vehicles that deliver product every day. In the past, knowing the drivers and having them work for us was important. Now with common driver management systems we know where those trucks are every minute of the day. We know exactly when orders are delivered because of bar code scans and proof-of-delivery technology. Owning or even controlling the truck is not important because all we really need to know is whether the order was delivered on time and complete. This lets us know who has the best cost, whether it is our truck or someone else’s. We have seen more investments in systems to manage inventory, processes, and supply chain transparency and visibility. This allows us to expand our breath of products and services. We are growing our SKU assortment and product mix and entering new product categories because systems can do a lot more work.
Do you have any additional comments you might like to share about your experiences in supply chain management?
The visibility of supply chains today is tremendous. We manage two supply chains: One is the physical product, and the other is information. The information can be more important than the product in some situations. If a retailer’s information system does not recognize a product that is shipped to them, the product will often be returned. But if the product is recognized by the computer, even if it was not ordered, the retailer will often accept it and even pay for it.
The ever-present challenge that supply chain professionals face is this: What’s next? Who or what will be introduced soon that will impact their supply chains? New technologies are being introduced almost weekly and can cause drastic changes in the flow of inventory and therefore supply chains. Supply chain managers must be aware of these coming changes and be prepared to embrace them.