The m-commerce fallacy

Mobile commerce may become a critical sales channel for your company, but before barreling in, consider its current drawbacks.

By Beth Schultz
Network World, 02/26/01

Have wireless Internet device, will buy.

That's the motto of the madly growing mobile-commerce industry. A slew of outsourcers, software makers and infrastructure vendors have sprung to life, convinced that anyone who has a cellphone, two-way pager or PDA is hankering to use it to buy stuff over the Internet. It doesn't matter that early users indicate they'd rather drive 20 miles in a blinding snowstorm than navigate through a transaction via a wireless device.

Clumsy user interfaces, cumbersome applications, low speeds, flaky connections and expensive services have soured many who have tried m-commerce, found a usability study conducted last fall in London by the Nielsen Norman Group. In the study, the consultancy gave 20 users cellphones that support the Wireless Application Protocol (WAP), a set of protocols designed to make Web content skinny enough to fit on cellphones and other small wireless devices. WAP was developed by the WAP Forum, an industry association in Mountain View, Calif.

The research firm had participants use the phones for one week and record their impressions. At the study's end, 70% of the participants said they would definitely not want to use a WAP phone again within one year. "WAP usability is horrible," says Jakob Nielsen, co-founder and principal of the Fremont, Calif., firm.

When m-commerce rules When and if wireless Internet access ever surpasses wired connectivity, "you don’t want to just be bolting on wireless," says Peter O’Kelly, senior analyst with market research firm Patricia Seybold Group. Click here for more...

Privacy and security concerns also dampen enthusiasm for m-commerce.

When it comes to privacy, location-aware applications, which use knowledge of a user's exact location, bother people most. Support for location-based services are often included with a hosting service or m-commerce infrastructure product. If not, it'll likely be available as an option. A commonly described scenario for location awareness is of a retailer zapping an ad to someone walking by a brick-and-mortar outlet. While some shoppers may be ecstatic to learn about a sweater sale 10 feet away, others would surely find this a creepy invasion of privacy.

Security concerns center on how to protect transactions as they pass from the wireless net, across the m-commerce infrastructure and into corporate back-end systems. The need for WAP and Palm translation gateways, typically housed within the carrier network, is the problem. As the gateway translates data from the wired to wireless format, a moment exists when that data is unencrypted. For a financial services application or other situation requiring ironclad security, those gateways should reside behind your firewall, says Mitch Bishop, vice president of marketing at e-commerce software infrastructure vendor Mobileum in Pleasanton, Calif. But at this point, he notes, most IT shops are comfortable keeping those gateways on the carrier premises.

M-commerce transformation IT's major task today is sorting through the mobile e-commerce mumbo-jumbo and making sense of emerging - and often conflicting - standards, technology and cultural issues. That's followed by deciding what applications are best suited for m-commerce and getting a handle on the development requirements and costs. Click here for more...

The other security concern is that the devices are so darn easy to lose and steal. A valid user ID and password could let the thief make purchases; server-resident extended user profiles often store credit card and other necessary validation information to avoid cumbersome data entry with the wireless device. M-commerce doesn't yet factor in support for security mechanisms such as digital certificates, although work has begun. For example, Entrust Technologies subsidiary Entrust.net is launching a trial WAP certificate service for developers and service providers building wireless Web applications.

Taking these concerns into account, it looks like good old e-commerce, the traditional way via a stationary PC and a dedicated Internet link, will serve many consumers' purposes for years. "Most consumers have little incentive to switch from e-commerce to m-commerce," says P.K. Kannan, associate director for the University of Maryland's Center for E-Service in College Park.

But even such unavoidable obstacles have done little to stall m-commerce maneuvering. Just days into 2001, for example, two wireless leaders - Sprint PCS on the service side and Palm on the device end - struck a deal through which Palm users will be able to access information on Sprint's wireless network. A co-branded version of the MyPalm portal will include Palm wireless, personal information management services, e-mail access, entertainment listings and m-commerce.

Vendors such as these are banking on statistics that paint a fantastic picture of a wireless world. Market research firm IDC predicts the number of m-commerce subscribers will grow from 399,000 in 2000 to 29 million by 2004. In the same time frame, IDC says, users will conduct $21 billion worth of m-commerce transactions.

The argument is that m-commerce is a natural byproduct of owning a wireless device. An AOL spokeswoman puts it this way: Of the company's 26 million subscribers, "Sixty-seven percent own a cellphone; 23% own a pager; 72% own a cell phone or a pager. We're really excited about the possibilities."

Of course, the logic is faulty. Just because someone owns a TV doesn't mean he patronizes a shopping channel. Just because someone owns an Internet-capable wireless device doesn't mean the person will ever shop with it.

Equally important is that the development effort required to conduct consumer-oriented m-commerce transactions is gargantuan.

You'll need to support myriad content formats, devices and mobile wireless networks. Some customers might be using WAP-enabled phones, while others use the i-Mode service popularized by NTT DoCoMo in Japan. Some devices devote much of their real estate to the screen; others can barely display three lines of text.

So nobody knows for sure whether m-commerce will grow as hardy as its e-commerce antecedent or will wither on the vine like the now-infamous push technology. Even the staunchest believers admit the talk far outpaces the reality. "I expect astounding things coming in the future, but we need a reset on the market projections and hype of today," says Peter O'Kelly, senior analyst with market research firm Patricia Seybold Group.

Nevertheless, nearly everyone agrees that m-commerce will materialize in some form. At its very least, it will become another must-have sales channel for the modern corporation.

The question for IT executives is how to proceed.

Most analysts say you need to begin researching m-commerce technology today with the goal of wireless-enabling at least a portion of your site within the next two to three years.

Kannan, who is also an associate professor of marketing with the University of Maryland's business school, says m-commerce will only become essential for companies whose products or services are time-constrained. In fact, he and others say, these companies should already have m-commerce initiatives in place. They include financial firms offering online trading, payment and portfolio management services, and travel (airline, hotel, rental car) and entertainment firms offering online booking.

Impulse sales have a moderate chance of success with m-commerce. Companies with strong brand names selling inexpensive products such as CDs, books, flowers and food can capitalize on location-aware applications. However, "only companies with the technological capabilities and having good experiences and reputations on the Internet can do this as a viable strategy," Kannan says.

In contrast, companies that sell expensive products with involved purchase decisions are not likely candidates for m-commerce. "Companies can use wireless to advertise and list products and store locations, but the chances that sales will be made through wireless orders is rather slim," Kannan says. He advises these companies to create a wireless channel by enlisting in wireless portals that help shoppers compare prices and identify closest store locations for purchasing.

Although the success of m-commerce is very much up in the air today, the argument most experts use for experimenting with it now is to be ready if the onslaught does hit. Seven years into the commercialized Internet, most companies still aren't in tune enough with Web processes to integrate them effectively with the existing infrastructure. "While it's too early for mobile commerce to become a profit-generating and commercially feasible service, companies must gain experience so if the mobile Internet takes off they'll be positioned to launch a great site," Nielsen says.

Consumer dining portal Food.com, which falls into Kannan's second category, buys into the get-started-now philosophy. It's working with Oracle ASP subsidiary OracleMobile on a hosted m-commerce application it expects to be ready for commercial deployment this spring. "We consider wireless-enabling our online ordering application a fairly major initiative in the grand scheme of things, but we're not tying any market projections or new market segments to it. It's just too early," says Adam Dubov, vice president of content and product management at the San Francisco firm. "We're really just looking to start the process and put the building blocks in place so we can unleash m-commerce heavily if warranted."

While dismissing ebullient market projections as overly optimistic is easy, if even a quarter of those predictions come true, wireless will be worth doing for most companies. And with its infrastructure requirements, it will never be effortless, or inexpensive.

Factoring in software, hardware, application development and integration, an m-commerce application used by about 50 people, built and maintained in-house, could easily run $500,000 to $1 million, says Mobileum's Bishop.

Mobileum's e-commerce infrastructure software package costs $100,000 alone. For that comparatively modest sum, you get all the experimentation and pilot projects you want. Move into production mode, and session fees kick in. A session is an open connection between a handheld device and the m-commerce platform. Mobileum offers two prices, one at 250 active sessions and the other at 1,000 active sessions. At the higher price, the base software costs $195,000 and enhanced wireless-service modules range from $95,000 to $125,000. Two developer tools are $50,000 apiece.

Whether you adopt m-commerce today or wait, with this much cash on the line you'll clearly need a strategy. Justifiably or not, m-commerce is already becoming a key differentiation among Web sites. Miss the m-commerce wave and you stand the chance of turning off once-loyal customers looking for a wireless destination.

Are you ready? Take our quick questionnaire to see what steps you should take.

Contact Signature Series Editor Beth Schultz at bschultz@nww.com.

Send this article to a colleague