The rampant failures of dot coms such as Pets.com and MotherNature.com brought general scorn and ridicule on the Internet flavor du jour. At a deeper level, however, the disappointment may be due to the fact that these companies were supposed to have worlds of advantages over their stodgy competition: the agility and will to incorporate one-to-one personalization into the buying experience from day one. This, we are meant to understand, guarantees long-term success by truly anticipating and fulfilling a customer's needs.
That it has demonstrably failed for so many could be seen as a black eye on the face of personalization. "I don't think it's really paid off yet, because it has been very much focused on pushing product and pushing prices," says Randy Covill, senior analyst with AMR Research. One of the pioneers of personalization, Edina, Minn.-based NetPerceptions says the game isn't over yet--the players have just been wrong. That's why the company has been downplaying its efforts in the pure Internet space, despite launching its career as the recommendations engine behind Amazon.com (which Amazon has since replaced with in-house software).
"It didn't look to us that dot coms were going to be the long-term winners," says Steve VanTassel, general manager for commerce solutions. His is a familiar refrain--the last 12 months were full of technology firms ditching dot com clientele for old-money brick-and-mortar establishments. NetPerceptions has the numbers to back up its claim to prescience, however. Dot com clients are still in the majority, but VanTassel says that greater than 75 percent of company revenues come from offline or "clicks and bricks" firms.
Whichever bandwagon the company may be on in its move to serve evolving old-economy names, one hot trend you won't hear NetPerceptions espousing any time soon is B2B. Although there are some business-to-business elements in the NetPerceptions suite, such as the ability for a company to offer its suppliers advanced visibility into end-customer demands and expectations, it turns out that consumer spending does still count in some circles. "There's just more money to be made here. [Retailers] spend millions on their decision support data warehouses, and if we can help them get more out of that, we believe they'll spend their money with us," says VanTassel. In order to run with the shopping mall giants, NetPerceptions is specializing in consumer retail solutions. "While you can analyze customer performance on a cross-industry basis, you can't evaluate product performance and profitability on a cross-industry basis. We have to survive in retail."
"It's backwards from traditional database marketing, where you start with a segment as a means of making the same prediction for everyone in the segment," says VanTassel. Instead of demographic data predefining a finite list of "buckets," affinity groups reveal themselves (with a little help from data analysis), and as tastes change and purchase histories grow, so does a company's understanding of the interplay between customer characteristics and buying patterns.
NetPerceptions has recently closed deals with Best Buy and Kmart, with the latter using the customer analysis suite to build its weekly product circulars down to layout and pricing by region. Macys.com is in the implementation phase in order to break out of its former static layout and cross-sell recommendations. "It was like stocking shelves--the merchandisers would say 'This is the featured product,'" says Macys.com director of technical research Gary Beaverman of the previous e-commerce system.
"In a department store, we have a great competitive advantage to be able to exploit a diverse product range and show we understand our customers better," he says--but only if the technology is in place. "We're up against direct mailers who understand targeted marketing typically better than we do, but have a more limited product range. If we get an understanding of direct sales while addressing customer needs in more diverse areas, we'll do very well."
Beaverman said that while Macys.com has no firm plans to do so, it has considered using NetPerceptions technology to create dynamic promotions, where special offers are made based on a customer's purchase history and interests (or lack thereof.) This is essentially "dynamic pricing" in a different guise, an extremely delicate topic. At a basic level, setting a dynamic price requires insight into the gross margin on a product, a customer's preferences and propensity to buy and the ability to make an offer in real time. VanTassel says that NetPerceptions meets all those requirements. "We could try to tailor pricing based on how much somebody liked something. But none of our customers are willing to do that, and frankly we don't know if it would end up being more profitable or less profitable for the vendor."
Covill thinks he knows. "I would not recommend to any business that they seriously try to push dynamic pricing," he says. "You're going to have to charge everyone pretty much the same price, unless you can justify a higher price in terms of some kind of service or value that you're giving one set of customers over another." In other words, simply knowing someone's place in line on the demand curve is not necessarily justifiable cause for a selective price hike.
Not that the payoffs from better targeting are anything to sniff at. NetPerceptions boasts of having helped one client boost its order rate on an e-mail marketing campaign more than twelvefold. The problem with congratulating any technology with short-term success is that there are any number of independent factors that may be working alongside a particularly impressive piece of personalization, from better offline marketing to a buying frenzy. "Over time, you should be able to see inventory turns go up, carrying costs go down, costs per customer go down, and you should be able to identify similarly performing customer segments," says VanTassel. "These are longer-term measures than most people think of."
Through both acquisition and partnership with firms such as Angara Converter, NetPerceptions is working to expand customer analysis beyond simple online clickstream history and into the vast realm of offline data integration--something Covill feels is important to the long-term success of personalization. "A lot of companies out there have a piece of the solution," he says, highlighting that in order to move to successful personalization, companies need more information to offer both the right product at the right price and a value-added service to retain the customer's attention. But database integration brings the risk of privacy backlash. "The problem is using data in ways that people didn't explicitly give their permission for," Covill notes. The wealth of information represented in such records as department store credit cards could give crucial insight that has been missing for some dot com retailers, if they can take advantage of the opportunity.