InformationWeek

ccessing the Web anytime, anywhere may seem like cyber-nirvana for many Internet junkies. But the wireless technology that enables such access has yet to successfully link retailers and shoppers in large numbers.

Several hurdles plague wireless operators, retailers, and potential users. Connection speeds are too slow for many users to start or complete a wireless transaction. Wireless applications still lack depth and breadth. Slow transmission rates and tiny screens mean that wireless shoppers don't get the graphical data, including retailers' cherished branding information. Amazon.com entered the wireless retail market in October 1999, followed by rival Barnes & Noble.com last May. But neither company will break out sales that come from wireless customers.

Nonetheless, the success of retailing and wireless, also known as M-commerce, is assured--in due time, says Craig Mathias, a principal at the Farpoint Group, a consulting firm. During the next two years, vendors and service providers will develop M-commerce products and platforms. But real growth and volume usage is at least four years away, he says. Research and consulting firm Accenture estimates that more than 130 million customers worldwide will spend $200 billion wirelessly between now and 2006.

Retailers are reluctant to embrace the wireless Web for various reasons, relating to both technology and marketing. "Wireless is great for McDonald's or a commodity reseller," says an E-commerce manager for an ultrahip clothing vendor who asked not to be identified. "But wireless may not be the right fit for retail clothing."

Another retailer is careful about where it devotes resources: Recreational Equipment Inc. in Sumner, Wash., was one of the first retailers to make its inventory available online--to wired users. That customer base is large enough for potential revenue in the medium to long term, says a Recreational spokeswoman. "We tend to stick to the least common denominator and concentrate on the bulk of our customers," she says, referring to those sitting at the end of fiber and copper connections.

One problem is that there's no universal "wireless Web," as some providers have suggested in their marketing material. Analysts agree that "wireless Web" is a marketing misnomer. In reality, AT&T Wireless, Sprint PCS, and other cellular operators each have their own versions of the wireless Web, but customers can't click from one operator's content to another's.

As it exists now, the wireless Web is a collection of the retailers, applications, and transaction clearinghouses that a particular wireless carrier has cobbled together.

"M-commerce hasn't really gone prime time yet," acknowledges a spokesman for AT&T Wireless. Although AT&T's retailers have seen sales increases, he says, "It's still just a blip on their radar screen."

Then there are marketing and technology hurdles. Applications exist in strange isolation from one another. Palm Inc.'s wireless application doesn't work with cellular PCS networks and terminals, for example. Or a carrier might introduce various wireless services with incompatible applications.

Further, the current generation of interfaces--cell phones or handheld computers equipped with either Cellular Digital Packet Data (CDPD) or Wireless Access Protocol (WAP)--remains clunky when it comes to browsing and buying online.

Keypads are awkward, even for basic functions such as sending E-mail. Plus, wireless modem connections lurch along at speeds of 9.6 Kbps to 19.2 Kbps--too slow to support the color, graphics, and sound that give the wired Web impact and vividness. Retailers that want their customers to browse and buy wirelessly must first dumb-down their content for text-only transmission.

Clearly, wireless is still very much seeking its niche beyond the well-established voice and paging markets. In the early '90s, groupware was supposed to be a killer application for wireless. That never materialized, Farpoint's Mathias says. Then came E-mail, which didn't need a lot of bandwidth and tolerated latency well.

But those characteristics aren't part of the wireless retail equation. Fast response times, whether to see a larger picture of a DVD player or get an online credit-card approval, are crucial. "When you're doing real-time shopping, you need a response now," Mathias says. "After a few milliseconds, the brain starts to think something's wrong." Wireless retailers must help customers reduce the time between knowing they have to do something and getting it done. "Customers want the instant gratification of solving a problem," Mathias says. "That's very powerful for most people."

Regardless of the medium, retailers want to deliver as much information as possible to their customers in this rich-media age. So if a big retailer spends $50 million to redesign its logo as part of a larger business effort, it's easy to understand why the company might ignore a medium it can't really exploit. Or if an online cosmetics seller differentiates itself by the number of colors it carries, why develop a sales channel that can't display those colors?

"Our Web site is chock-full of graphics-intensive material, so we're probably not going to partner or reformat the site to go with mobile commerce," says a spokesman for Circuit City Inc. in Richmond, Va. "We're not sure the type of product we have is something that would work in an M-commerce setting."

It's a common concern, especially because brand name and brand equity are the lingua franca of retail, says Randy Covill, senior analyst for retail applications services at AMR Research. Retailers recoil when shown the tiny screen of a cell phone or handheld computer that contains only product information and pricing, without logos or catchy graphics that lure people to their sites. "It commoditizes the product and strips away all their brand messaging," Covill says. "They're scared to death that customers will use these devices to do price comparisons."

If wireless terminals are to operate faster and be able to deliver the graphical files retailers demand, the network that carries the traffic must operate faster. The artificial bottleneck that keeps speeds so low is a function of the radio spectrum used by wireless operators.

Those operators uniformly point to third-generation, or 3G, wireless systems and their 64-Kbps and higher throughput that will make a big difference in the poor showing of today's M-commerce.

Operators and terminal vendors herald 3G as the fix for many of wireless data's pitfalls. Color, graphics, and sound will bring the wireless browsing experience in line with what customers can get only at their desktops. Time spent waiting for a page to load or for password approval will be significantly reduced. What's more, customers could move seamlessly between fixed and wireless networks with their computer and form factor of choice.

But such wireless good times are probably a couple of years away. Sprint PCS is preparing to have 3G service available by the end of next year and claims it will be the first American operator out of the chute with 3G broadband wireless services of 144 Kbps and higher. AT&T Wireless promises limited availability next year and a nationwide rollout in 2003. Meanwhile, operators such as Cingular Wireless and Verizon Wireless also are working to launch 3G networks and services in similar time frames.

Browsing, buying, and getting credit-card approvals are the easy part where wireless retail-application development is concerned. A bigger challenge is overcoming the fragmentation between Palm's Web Clipping application and the CDPD-or WAP-enabled apps that mobile phones use. Some wireless retailers, such as Barnes & Noble.com, split their bets and make sure both kinds of wireless users can access their goods.

Fragmentation also exists within a single operator's own portfolio of services. AT&T Wireless already has its PocketNet service for wireless data transactions. It's also working with Qpass Inc. for a secure digital-wallet capability, and with Infospace Inc. for electronic coupons that can be used with wireless retailers.

AT&T is testing those capabilities, and they should be available by September, the carrier says. The company also is working with NTT's mobile communications subsidiary DoCoMo and its iMode wireless data service, which has been phenomenally successful in Japan.

Not all those applications or system types will work together. But AT&T doesn't think it's further fragmenting the market or impeding the development of a wireless retail mass market. Rather, AT&T Wireless says it's important to cover as many spots on the board as possible.

"This isn't at the scale of Apple, Linux, and Microsoft, but it may become more of an issue as people buy more devices," the AT&T spokesman says. Of the company's 15 million customers, only 500,000 have access to its wireless Web services.

Still, fragmentation plagues every sector of the wireless retail market--operators, terminal vendors, applications developers, would-be retailers, and customers.

By all accounts, the upside of wireless is huge, even if the path there is unmapped. Operators must upgrade to accommodate higher throughput and richer applications. Terminal vendors may want to rethink the number of standards and platforms for which they develop products. And customers have to be convinced it's easy, fast, and secure.

That's an enormous balancing act, Farpoint's Mathias says: "Retailers can't count on people coming into the store or calling up and getting put on hold. People want to buy things and communicate with those that sell to them. The easier we can make that, the better."