It has been almost five years since Jerry McCabe's presentation at the Aftermarket eForum drove home one of the most important messages the aftermarket had heard in many years. The presentation I refer to was about the importance of data in an increasingly computer dependent aftermarket. It was in that speech that McCabe uttered the immortal words that have become a mantra for technologists in the aftermarket: "It's all about the data."
The essence of McCabe's presentation was that the only way to take full advantage of all available electronic and mechanical technology is to have perfect data and to keep it perfectly synchronized throughout systems, both internally and throughout the entire supply chain. It seems like a simple enough idea, yet five years later the aftermarket still is struggling with this very issue. Recently, while attending a couple of industry meetings, I was reminded of how prophetic McCabe's words were.The first instance occurred at a meeting of the data receivers group, which is comprised of all the major retailers, program groups and commercial electronic catalog (eCat) providers. The group came into existence to deal with the frustrations and problems that these companies are experiencing with supplier data. In a nutshell, the group contends that the majority of aftermarket companies have data that is in some way corrupt or incomplete, creating a situation where a significant percentage of the parts they offer cannot be correctly identified for sale.
That's a statement that causes top executives at most aftermarket suppliers to recoil in disbelief. They tend to assume that because their data is in an industry-standardized format it must be accurate. They don't understand that standards-compliant doesn't necessarily mean it is right. To have "perfect data," manufacturers must go back and retool their processes to manage data for electronic use. "Perfect data" is data that is accurate first and standardized second. Data standards are not a substitute for effective data management processes or for accurate data.
That is not to imply that standards are not important; it is an issue of priority. Standardized data is essential for data synchronization. This brings me to the second instance that caused me to flash back to McCabe's speech.
That occurred at a category management meeting. The group of assembled suppliers and resellers were lamenting the lag between the time when they provide point of sale data to The NPD Group (the service provider for category management sales data) and the time NPD gets it aggregated and reports back. As one committee member put it, "Why are we not able to view NPD data for March sales until July? A lag that long makes the data much less useful."The reason for the delay is that NPD has to resolve data errors manually. In the normal process, retailers and distributors send records of their actual retail sales information to NPD. If NPD can match the resellers records to "known good part numbers," the transaction is recorded in their data file. Each month, NPD must resolve tens of thousands of exceptions in the data stream it receives from resellers.
That means a person must open each record and interpret what part was sold. The problem typically is subtle differences in data fields, compressed part numbers, reseller-added prefixes or even dropping the lead zero from a UPC code. But because computers take things so literally, any variation in data, however trivial, results in an error that requires a person to "interpret" the data and fix it. As fast as they build a map to fix one error (the process of writing a formula that tells a computer that a PF8A is the same as a PF8), another one occurs in a different product category for a different reseller. NPD every month processes around 20 million sales records provided by dozens of resellers. Just a 1 percent error rate results in 200,000 records that have to be individually inspected and interpreted. The problem is the errors keep occurring, and patches in the form of data maps don't permanently fix the situation. That means human intervention is required over and over, adding time and cost to the process.
The real key to McCabe's message is the second part: "The only way to take full advantage of the electronic and mechanical technology available today is to have perfect data, and to keep it perfectly synchronized throughout systems."
Data synchronization is nothing more that the process of matching or aligning the data stored in different systems. The data creator (the supplier) and the data receiver (the reseller) both send files of the data they have on a particular line to a third party where they are evaluated for discrepancies. The differences are flagged, reviewed by both parties and adjusted to the mutual agreement of the trading partners, bringing the two data sets into synchronization. The focus in the exercise is typically on attribute data (part numbers, dimensions, description, prices, etc.), although increasingly it is becoming apparent that data synchronization will resolve many problems on the application side.
This brings me to the two instances I mentioned. The data receivers group and the category management committee are perfect examples of how the application of McCabe's advice will fix two major industry problems.
The problems the data receivers group experiences could be resolved with comprehensive data synchronization between manufacturers and eCat providers (be they resellers or third parties). If every leading eCat would send a list of the parts they have in their catalogs and what applications they fit to a third party to be synced with the manufacturers' data files, the coverage problems and application errors we suffer in the aftermarket could be virtually eliminated.
This would require unprecedented cooperation and collaboration between competitors since there is no "industry standard" eCat with which they can synchronize their files. However, one could speculate that such a bold move might be in the best interest of eCat providers, because it might thwart the growing momentum to create an industry standard eCat. But that is a discussion for another time.
The category management committee is on a much less slippery slope. At its recent meeting, it reviewed the state of data confusion that exists between suppliers, retailers and their data aggregation service, and concluded that a data synchronization exercise just might be the best way to resolve the delays they have been experiencing. Happily, that has led a group of progressive manufacturers and resellers to volunteer to engage in data synchronization on their category management data with NPD, their third party data service provider.
The group agreed to a three-way sync of critical category management data, including part numbers, bar code, brand ID code, AAIA product management category code, part description and the AAIA part type identifier.
Such an exercise holds the potential to radically reduce, if not eliminate, data errors in reporting and will speed up the turnaround of the category management reports on the lines that are synchronized. Clean data synchronized between manufacturers, resellers and the data service (NPD) would mean the category management reports could be produced in days or weeks instead of months. That is the sort of collaboration and use of technology that will help sell more products and produce a better ROI. The category management committee will monitor the results of the data sync, and if the test program demonstrates the anticipated success, a broader project will be initiated.
The powerful observations made by McCabe five years ago summarize succinctly the three steps the aftermarket must take to realize the full benefits of technology and automation. The first two steps involve perfecting the data and standardizing it. That means the "owners" of the data, the manufacturer or supplier of the product, must systematically compile full and complete information about their products, including attribute data, numerical identification, dimensions, images, prices, etc., as well as application data. That information next must be tediously checked for absolute 100 percent accuracy. Then, and only then, it must be stored in industry standard data formats (ACES and PIES). Finally, a robust data synchronization process will spread accurate, standardized data throughout the aftermarket.
Broader implementation of all of this is essential if the aftermarket is to realize the promise of technology and automation. Until manufacturers commit the time and resources to create "perfect data" that is complete, accurate, compiled into industry standard formats and resellers agree to synchronize their files with those of their suppliers, we will continue to flounder in our efforts to automate. The new signs I see are encouraging, as necessity is bringing many influential groups to the realization that it's still all about the data.
Bob Moore is president of Bob Moore & Partners, a consulting firm that specializes in the automotive aftermarket. Moore can be reached at [email protected]
