SEMA Business Technology

By Bob Moore & Dan Jondron

Why Data Matters
If Data Is the “Fuel” of Computers, You Want the Good Stuff

Sometimes when those of us who are less than computer geeks read about supply-chain new technology, it seems that the people writing about the subject are obsessed with “data.” The articles all seem to go on endlessly talking about the importance of full and complete data or about data standards with cryptic acronyms like PIES and ACES. Why is data so damned important?

Data is the fuel that makes computers work. And just like contaminated fuel in an engine, if you put incomplete, inaccurate or otherwise contaminated data into a computer system, it will sputter, spit and stall.

To understand the importance of data, one must consider how computers “think.” Computers don’t actually think like people; they just process data. And they do that very literally. It’s one very good reason why computers don’t drive cars. Tell a computer to go straight for 2.2 miles and then turn left by the gas station, and a computer would go straight, all right—straight off the road at the first slight curve. When the computer encounters a Circle K when it is time to turn left, it might not recognize a convenience store that sells gas as a proper gas station and it would just stop and generate an error report.

While computers can process a lot of data very quickly and very accurately, they lack the ability to reason and interpret the way the amazing little computer we call a human brain can. And that is why data standards become so important for aftermarket trading partners that want their computers to interact with each other.

Data variations and data formats are endless. Even those issues that seem intuitive or obvious to our “human computer” are not so obvious to computers. Consider a few examples that tax a computer’s ability to “understand” data:

Is a pound, lbs. or #?—If a manufacturer sends product information to a distributor that describes the shipping weight of a carburetor as 6 lbs., it’s pretty obvious that the product weighs six pounds. However, the distributor’s computer might refer to that weight as 6 # or 6 pds, not 6 lbs. Unless there is previous agreement to describe a “pound” as “lbs.,” one computer will have no idea what the other is referring to. God forbid that the customer is in Canada and his system is based on metric weights. PG=pounds gross, PN=pounds net.

What’s a half an inch?—Is the carton 6.5 inches tall or 6.5" tall? Note that there are two variables in this example. First is the expression of one-half as either a fraction or a decimal. Computers don’t do well with fractions, as they “see” slashes as markers between expressions (as in http://yahoo.com/mail). The other issue is the abbreviation for inches. While our human computer easily translates (") as inches, a computer is much more comfortable with the PIES standard IN.

Gimme a case of those—Packaging is another area open for interpretation, this time even to the human computer. It seems there is no convention for common terms such as cases or cartons. One man’s case is another man’s carton. Then there are oddities such as inner-packs or self-merchandisers. Again, the PIES standard brings definition for all these issues.

You say Chevrolet, I say Chevy—As intuitive as this one seems, any database will sort Chevy parts into a separate category from Chevrolet parts.

How many of what?—In the aftermarket we use a lot of descriptions to define “how many,” or units of measure. When someone asks for one, did he mean one kit, one set, one pair, one carton, one foot, roll, bulk, quart, gallon, case, box or just plain “one.” Think about this: If braided fuel hose is purchased in 50-foot rolls and sold in one-foot increments, what happens when you want 50 feet but the supplier’s system sees the order as rolls? Somebody is making a large return.

To avoid such “data disasters,” our industry trade associations—including SEMA—have created standards to make computer-to-computer communication possible between trading partners. The standards quite literally standardize the way the industry describes various attributes of our products and what they fit. These standards are easy to implement. In many cases, compliance is just a matter of changing column headings or the abbreviation that is used. In other instances, it may require adding a field or modifying one. The changes are usually small, but they can have a big impact once implemented. A good example is the “short product description” field in PIES. Very often, manufacturers will put their part number or an internal description of the product in that field. But consider it from the customer’s perspective: Have you ever pulled out a cash register receipt that had a long string of meaningless letters and numbers that cost $19.99 and gives you no clue what it was you bought? At a time like that, you will appreciate the manufacturer that populated the “short product description” with the words “air filter.”

Many of the same articles that are “data obsessed” talk about how incorrect and inaccurate data is costing the specialty-equipment market billions of dollars. For most of us, that seems improbable at first, but as more is learned about data and how unstandardized data contributes to more computer errors, those projections of losses don’t seem so outlandish.

There is a story told by one of the senior buyers at W.W. Grainger, the large industrial distributor, about data that is worth sharing. The story is set in the late 1990s around the time that Grainger was seriously migrating its business from paper catalogs and telephone ordering to a web-based eCat and online ordering. It seems that this buyer had a great vendor, one that did just about everything right. The product was great, the service fantastic, but the supplier wasn’t very keen on addressing Grainger’s endless requests to provide full electronic data on the products being provided. This supplier took the attitude that it didn’t much care about all this computer hocus pocus and, besides, it knew it was a great manufacturer and supplier (having won dozens of preferred-supplier awards). At first, Grainger did the data work for this great supplier but eventually gave it the ultimatum: Get us the product data we need, or we will find another vendor who can, even if its product isn’t as good. The supplier called Grainger’s bluff and lost. The line was replaced with one that wasn’t quite as good but could provide the product data Grainger needed to fuel its eBusiness needs. That might not happen today in the specialty aftermarket, but the day is coming…and soon.

Getting your data standards compliant also greatly expands the potential size of the market and channels into which both manufacturers and distributors can sell. Major retailer chains and new-car dealers challenged by shrinking margins in their core product areas have determined that selling performance and accessories products is an important strategy for expanded profitability. Since these businesses will do most of their transactions through special orders, they will need to be able to source, check stock and place orders while the consumer is standing in front of them with their checkbook or credit card. A transaction of this nature will only occur when our computers can talk to each other and exchange complete product and application data in a standardized format.

There is an extensive feature on this issue entitled “Cutting Cost vs. Selling More" that examines the upside of expanded selling opportunities along with the efficiency that technology offers.

Taking advantage of the opportunities that technology offers requires the existence of clean, standardized data “fuel” to power your computer. Realizing those benefits, whether through lower operating costs or the ability to sell more to new and expanded markets, requires SEMA manufacturers to get serious about getting their data act together and sharing it with their customers. (Manufacturers in particular should read the article entitled “Manufacturer Product Data Services" to get more insight into preparing and sharing their data.)