November 2001

 
mCommerce -- DOA? Or A-OK?

BY MARK E. McDOWELL

The hangover has been painful. WAP failed to deliver on its promise to make the desktop available on the mobile device. Beset by painstakingly slow access and nested menus reminiscent of DOS days, WAP has become persona non grata among North American and European wireless consumers. Compounding the problem has been the failure of carriers to deploy location technology within the expected timetable. One month shy of the October 1, 2001 E911 Phase II deadline, every major carrier in the U.S. has requested waivers or extensions. The Public Safety Answering Point of San Francisco has conspicuously announced that none of the carriers in its region are able to provide the level of accuracy required by the FCC mandate. To make matters worse, one of the most promising publicly traded firms in the location technology industry, US Wireless, Inc., announced in late August that it would seek bankruptcy protection under Chapter 11.

With WAP far from consumer consciousness and location technology beyond the horizon, is mobile commerce dead on arrival? Yes and no. Mobile commerce will probably never see $200 billion under the original paradigm where subscribers use their phones to go shopping on the Web. On the other hand, mobile commerce may be resurrected under a different paradigm -- one in which retailers have the ability to send targeted ads and coupons to willing subscribers, not using WAP, but rather using simple text messaging (short message service, or SMS) and perhaps someday, wireless instant messaging.

You Can't Go Window Shopping With A Cell Phone
If e-commerce is a global shopping center, then mobile commerce is a corner convenience store. Early returns from Japan, and to a lesser extent Europe, have shown that mobile commerce is well-suited to inexpensive, consumable items: Ring tones, animated figures, virtual girlfriends, parking meter payments, and sodas. Simply put, mobile commerce today is superb for impulse purchases.

And yet, ironically, the WAP experience is anything but impulsive. To conduct a simple mobile commerce transaction, a wireless subscriber must: (1) Have a WAP-enabled cell phone with the WAP service activated, (2) Place the phone into a WAP session and explicitly agree to pay a fee, (3) Enter a URL using a torturous keypad entry scheme, or, if the subscriber is lucky, thumb through several layers of nested menus and "next" softkeys to find a bookmarked URL, and (4) Navigate through the destination WAP site to make a purchase, and on and on. Even the stalwart early adopter may lose patience, and that's before the credit card transaction, since there is still no "wireless wallet." The impulse to buy is suffocated.

In this paradigm of mobile commerce, who is missing? The seller! The most savvy retailers will build WAP sites and register themselves in wireless search engines, then wait passively for a subscriber to browse by. The vast majority of retailers don't know what a WAP is or that they're missing one. And yet, paradoxically, even the most low-tech retailer would love to attract wireless phone users, who increasingly represent the wealthiest half (or more) of the population. Under the WAP paradigm, retailers are relegated to passive status, and despite their desire to sell goods and services and their willingness to pay to promote themselves, they have no way to access the consumer.

In order to sell, retailers need to be able to reach consumers and communicate their value propositions. Merchants and the media have evolved together in lockstep for a thousand years through print, radio, television, and the Web. Today, merchants need a way to advertise on wireless phones, especially to consumers who are willing to receive their messages. The wireless Internet paradigm of today -- WAP -- does not fulfill this basic economic need.

The Dual Myth of E911 and Location Accuracy
The armchair quarterbacks of the wireless Internet reasoned as follows: Carriers will deploy new location technology in order to comply with the FCC's E911 mandate for 100-meter caller location accuracy; the same technology will be used by retailers to bombard consumers with coupons when they are nearby; and, retailers (for some reason, Starbucks is always the marketing example) will profit enormously, sharing their wealth with the wireless carriers.

This vision is fraught with problems. How does the retailer know the consumer's mobile phone number? (Does Starbucks keep a list of names of every customer?) How does the retailer actually get the location information? How does the message actually get to the consumer? Must the consumer enter a WAP session and browse to the retailer's site in order to see the coupon? What happens if, heaven forbid, your office is located next to a Starbucks? Does your phone beep all day long?

Various companies have engineered clever answers to most of these questions. Their answers call for investments in IT infrastructure that are formidable in complexity and cost. Some carriers are even commencing with these solutions, determined to realize a "dual use" advantage from their E911 investments and they froth over mobile commerce revenue projections.

Now step back. Rather than engineering clever workarounds to the mechanical challenges of mobile commerce, let's question the fundamental assumptions. Should the same technology employed for E911 be used for commercial location services? And, even more fundamental, how important is location to mobile commerce?

The system that handles emergency 911 calls, the Mobile Positioning Center (MPC), has a relatively simple job, which it must do with high reliability. The mobile switching center directs 911 calls to the MPC, along with the identity of the base station, which is hosting the emergency call. The MPC routes the call to the Public Safety Answering Point nearest to the base station, along with the cell phone number and the location of the base station. Emergency response vehicles must comb the coverage area of the base station, hopefully while speaking with the caller on his or her cell phone. As 911 gives way to Enhanced 911, the accuracy of the caller's location improves dramatically, from a radius of several miles to a radius of about 100 meters. Nevertheless, the basic call processing and response remain unchanged.

While emergency call processing requires that the MPC must do a basic job over and over again without variation, like the sorcerer's apprentice, commercial location services place quite different demands on a MPC. Notably, the MPC must provide an interface to the Internet and dozens (or hundreds) of software applications, which make use of location information. The MPC must geocode, or translate, geographic information into street addresses, postal codes, and neighborhood descriptors, and the MPC must respond to inquiries outside the wireless network rather than merely reacting to handset-initiated inquiries. The unique demands of commercial location services will only grow, soon adding privacy controls of personal location data, rich billing system interfaces, and much more.

Left unchecked, the MPC of the near future will become a "Bride-of-Frankenstein," cobbling together the essential functions of emergency call processing with the complex and ever-changing needs of a commercial location server. The solution is to decouple emergency call processing and location services processing. Let the label "MPC" continue to define the emergency call processing system and assign the new label of "Location Server" to the commercial location services system. This simple act of separation removes frighteningly high processing loads from the emergency call processing system, while at the same time giving commercial location services leg room to grow as the industry evolves. More pragmatically, the dependence of mobile commerce on E911 compliance is broken -- the two can proceed with a high degree of independence according to separate regulatory and market demands.

With the Location Server liberated from the MPC, we can now address the issue of location accuracy requirements. Do commercial location services require 100-meter accuracy? In the vast majority of cases, no. The earliest concoctions of mobile commerce showcased parlor tricks in which Starbucks sent a "$1 Off" coupon to the unsuspecting wireless subscriber just as she rounded the corner near the store. Why was this level of accuracy written into the script? Because it was possible with an MPC. But such accuracy may not be necessary at all. Simply being in the same ZIP code as a Starbucks is probably enough to merit sending an ad. For that matter, Starbucks is everywhere -- why does it matter where the subscriber is? Just send a coupon and let the consumer go to her favorite Starbucks.

The question of location accuracy is not purely academic. There are enormous capital consequences for wireless carriers. Today's digital wireless networks (2G) already know the identity of the base station and the orientation of the antenna that is handling a call, information that can be readily translated into a ZIP code. Going from ZIP code-level accuracy to city block level accuracy entails a massive expenditure of funds, either on network-based location technology or on GPS-enabled handsets.

A modest proposal: Carriers should completely decouple E911 compliance from commercial location services. Location Servers, independent from MPCs, should provide ZIP code-level location data to retailers and other commercial software applications. This can be done today with minimal capital outlay and will serve the vast majority of mobile commerce needs, which are impulsive purchases.

The Answer Was Under Our Noses All Along
If in this gedanken experiment* we have broken the stranglehold of E911 over mobile commerce, what about the equally tight grip of WAP?

One needs to look no further than the typical junior high school for the answer. In the U.S., teenagers will tell you that Internet instant messaging is indispensable; in Europe and Asia, they will say the same about SMS. If these two global, and completely unplanned, phenomena could be merged into wireless instant messaging, the grip of WAP would be broken. SMS provides a convenient and scalable mechanism for retailers to push their ads and coupons to wireless phones. Instant messaging, through the all-important buddy list, is the vehicle for a user to specify his or her availability and to grant permission to specific vendors to send messages. The buddy list becomes, in effect, the dashboard by which consumers publish their availability and interests.

Of course mobile commerce is not dead! Mobile commerce has been crippled by our overzealous drive to make it conform to the shape of E911 and WAP. We have tried to impose particular technical solutions on top of impulsive human behavior, simply because the technology was present. Mobile commerce does not require the accuracy of E911 call processing; it may not require much location data at all! Nor does mobile commerce require WAP, the grandly planned flop, when in fact the brilliantly accidental IM and SMS will do quite nicely. c

* A quick visit to www.dictionary.com yields the following definition:

Gedanken is a German word for thought. A thought experiment is one you carry out in your head. In physics, the term "gedanken experiment" is used to refer to an experiment that is impractical to carry out, but useful to consider because it can be reasoned about theoretically. (A classic gedanken experiment of relativity theory involves thinking about a man in an elevator accelerating through space.)

Mark McDowell is the president and COO of Invertix Corporation, a privately held company headquartered in Alexandria, VA, with operations in North America and Europe. Invertix's mission is to enable mobile network operators to create revenue and customer loyalty by providing the world's most powerful end-to-end multimedia messaging solution.

[ Return To The November 2001 Table Of Contents ]