Manufacturers are finally innovating the service chain, using new analytical software tools to cut warranty costs and identify product failures faster.
John Kerr wants to eliminate the service call.
The general manager for global quality systems at $14 billion appliance manufacturing giant Whirlpool Corp. (Benton Harbor, MI), Kerr is in charge of the company's sweeping effort to organize the data contained in thousands of warranty reports, field service repair notes, and dealer and retailer feedback forms in a database that can be mined and analyzed.
The knowledge gained from such systems, Whirlpool believes, can then be used to pinpoint component problems, identify patterns in quality problems, and set off early warnings of possible product failure. And that can allow Whirlpool and Kerr to reduce -- if not eliminate -- service calls.
Automated warranty management and analytic systems like the one Kerr is working on can condense the time between identifying a problem and fixing it in the supply chain. In the past, that process could take six months or longer, Kerr says. The introduction of better tools that can gather and analyze post-sales data means cutting this lag time significantly, to the point where a design change or component replacement can be made before thousands of potentially defective appliances are shipped.
And avoiding that kind of mistake can save big money. Just ask Dell Computer, which, analysts say, is spending upwards of $400 million to replace 4.1 million defective laptop batteries.
Warranty management has become the latest area of innovation for appliance manufacturers such as Whirlpool, which uses a warranty management solution from SAS Institute Inc. (Cary, NC), as well as for manufacturers of automobiles, heavy trucks, and consumer electronics.
In the automotive industry, these initiatives were sparked by the Transportation Recall, Enhancement, Accountability, and Documentation (TREAD) Act, a legislative response in 2000 to a string of fatal tire failures on Ford Motor Co. sports utility vehicles. The law requires auto manufacturers to record and track all data received from warranty reports, dealers, and repair shops, with the aim of identifying and fixing potential manufacturing defects before they become widespread in the market.
TREAD was followed in 2004 by the Sarbanes-Oxley Act, which, in addition to forcing companies to document financial controls, required manufacturers to break out the cost of things like warranties and post-sales service.
Combined, the two laws have opened manufacturers' eyes to the true cost of warranty management. For example, GM and Ford together spent $8.7 billion on 2005 warranty claims, up from $8.3 billion in 2004 and $8 billion in 2003, according to a study by Warranty Week, an industry newsletter, which looked at data reported to the Securities and Exchange Commission. The top 50 U.S.-based warranty providers together reported $22.2 billion in warranty claims last year, up from $20.4 billion in 2004 and $19.2 billion in 2003, according to the same report.
Such numbers have forced managers to revisit their service strategies and explore opportunities for change. Warranty management "is one of the biggest areas of controllable expense," says Thomas Wright, industry principal for automotive at SAP Americas (Newton Square, PA).
"Coin-operated process"
Warranties are akin to insurance policies. Offered free or for an additional charge, the policy covers the cost of repair should something go wrong with a product. Since repair is either free or discounted to the consumer, when something needs to be fixed, the manufacturer reimburses the repair shop.
As with any other claim, the repair shop must document the problem on a form. Generally this is done through coding -- e.g., 0435 might mean "defective belt." In addition, there might be a space on the form for comments, where the repair technician might write, "Belt was worn due to friction."
But, historically, there have been problems with the data generated by such processes. For one thing, warranty and repair data tends to be unreliable, says John Thomann, a private consultant in the areas of field service, service parts logistics, and warranty management.
"Data tends to be scurrilous," he says. "The repair man wants to be reimbursed, so he will code a report in such a way that he will [be]. It's a coin-operated process." Legacy warranty management systems didn't always provide incentive for accurate or complete reporting by service providers, who often feared payment would be held up or denied.
The process also lent itself to error. Oftentimes, manufacturers would demand that repair shops deal with hundred of codes. "No one can possibly know all of them, or have the time to look through the whole list to find the precise code," Thomann says.
Finally, information hasn't been regularly shared between manufacturers and their suppliers, mostly out of fear of finger pointing, which trumped improving quality control, Thomann says.
Now, armed with a better understanding of the cost of quality, manufacturers are realizing they can use IT systems to mount a double-sided attack on the cost of warranty management and post-sales service. First, they can build more accurate databases that lower the cost of warranty fulfillment by verifying that reported problems are eligible for coverage and identifying and eliminating fraud. Second, they can use analytics to track post-sales service complaints and nip product problems in the bud.
The warranty management systems at the heart of these efforts manage, transmit, and store documents and create the historical record. Warranty management functions can be spread out among different systems from different vendors, beginning with enterprise resource planning (ERP) systems and customer relationship management (CRM) systems and extending to more specialized warranty systems.
SAP, Oracle Corp. (along with recently acquired Siebel Systems), and Infor Global are among the providers of ERP and CRM platforms that support warranty management, claims processing, information gathering, and storage.
Manufacturers like Whirlpool are attempting to take all the data collected by warranty management systems and create analytics that can reduce the time needed to take corrective action. Such analytics mine the record created by warranty management systems, looking for patterns and trends that can be used to identify defects, forecast lifecycles, reduce risk, and improve design.
SAS Institute Inc., Cognos Inc. (Ottawa, ON), Business Objects SA (San Jose, CA, and Paris), and Oracle are some of the vendors that provide business intelligence software that uses data from CRM and warranty management systems to identify and analyze problems, support decision-making, and help spot trends.
Attensity Corp. (Palo Alto, CA) -- another Whirlpool vendor -- as well as SPSS Inc. (Chicago) and ServiceBench Inc. (Fairfax, VA) provide analytic tools that specifically look at unstructured data contained in notes, reports, memos, and other warranty-related documents.
Three Phases
So where should manufacturers focus their efforts to innovate the warranty management process? SAP's Wright divides the reengineering process into three phases. The first is an overhaul of transactional processing -- the mechanism by which manufacturers deal with their service networks. This means streamlining codes and standardizing the reporting process so that manufacturers get more accurate information and dealers and service shops get paid more promptly.
The second phase is supply chain integration. This pushes some of the cost of quality back to suppliers. This, Wright says, is where trend analysis enters the picture. "It's here where I ask: What component is failing? Is it single source or not? Is it a product flaw or spec flaw?"
The automotive industry is in the middle of a big transition to this type of analysis. Fifty percent of the systems in an automobile are electronic, Wright says. With them come all sorts of operational parameters and failure profiles. An otherwise sound component, for example, may fail under a certain combination of conditions -- internal engine temperature, outdoor humidity, state of battery charge, amount of vehicle vibration.
Warranty analytics can correlate variables from widespread sources and extract patterns that would have gone unseen in the past. The most potent example was the Ford-Firestone tire failures in the late 1990s. In those cases it was a combination of air temperature and vehicle speed that was causing tire blowouts. Over the course of several years, incidents were reported from Arizona to Saudi Arabia, but the underlying cause remained elusive. Today's analytics, say executives such as Michelle de Haaff, vice president of products and marketing at Attensity, might have pegged the problem sooner.
Attensity is part of the third phase of Wright's reengineering model -- analysis of unstructured data, including the written comments, emails, and repair notes that account for 85% of all post-sales reports. An SAS partner in the Whirlpool implementation, Attensity offers a text analytics tool that captures unstructured data and uses it to build matrices that can be easily searched and mined for predictive information (see sidebar). "Manufacturers get early insight. A response in 200 days is cut to 100 days," according to de Haaff.
All Your Parts
Predictive analysis may be the high-tech side of warranty and post-sales management, but that still leaves many nuts-and-bolts issues -- literally. Parts supply is another aspect of post-sales support, and warranty management systems can be used to create the supply models for warehouses, depots, and distribution centers, says Bob Salvucci, president and CEO of MCA Solutions Inc. (Philadelphia), a supplier of service planning and optimization software.
A key question for any company, Salvucci says, is where it should place inventory in the supply chain in order to satisfy service contracts. Boeing, an MCA customer, has 250,000 active parts in more than 1,000 locations, Salvucci says, all designed to be delivered within 24 hours.
When parts inventory management systems like those from MCA are tied to post-sales data, manufacturers are in a much better position to set up distribution channels that allow parts and services to reach customers more quickly.
And post-sales service support can be just as important to today's virtual manufacturers such as MCA customer Cisco Systems as it is to Boeing and Whirlpool.
With manufacturing increasingly outsourced, post-sales service becomes a chief element of brand equity, Salvucci says. For instance, when it comes to its routers, all Cisco owns is the intellectual property. A serious defect in any area of the supply chain could cause serious problems that would ultimately reflect on the brand. The key, then, is to innovate and stay as informed as possible and be able to act quickly on information.
"Cisco doesn't do manufacturing, it doesn't do field service, it doesn't do repair," Salvucci says. "It's a virtual company. The brand is invested in the customer relationship. And that relationship is managed with warranty and service contracts."
Steven Titch is a freelancer based in Houston.