The M-Commerce Paradox: If You Succeed, You'll Fail

After we ran a story in the last issue about some Mobile-Commerce experiments at Macy's and Best Buy, one retail exec at a very large chain who has worked extensively with mobile expressed skepticism that some of these projects would ever work at full scale.

"I wonder when people will realize that mobile devices communicate via, ummm, radio?" asked the IT exec. "And that microwave radio signals (which GPS, mobile phones and microwave ovens all use) don't transit solid surfaces, especially conductive ones like metal mall roofs, all that well? And that carriers make no promises relative to in-building coverage (and virtually no promises relative to out-of-building coverage)? And the U.S. government makes no promises at all relative to GPS signal penetration or even availability?"

His comments were particularly based on retail efforts to use GPS to identify customer locations. Prodded by our reader (nobody—and we mean nobody--can be more cynical and snide than us and get away with it), we ventured into some area malls with an older GPS unit. And, as predicted, the mall's roof trumped the military satellite.

But the M-Commerce paradox isn't limited to GPS and malls. What if a major retailer in, let's say, New York City, decides to do a mobile payment trial? To give it a meaningful chance, the retailer throws every marketing tool at it and offers steep discounts to anyone who uses the new mobile app on a particular day.

If the marketing campaign is successful, thousands of consumers will flood into the store that day to try the app and get the discount. What's likely to happen then? A huge amount of nothing, courtesy of bandwidth unable to handle that traffic in such a compressed time period.

Actually, it's not fair to say nothing will happen. More brand damage will be done to that app than its owners could have envisioned in their worst nightmares.

Consumers will try the app and it simply won't work. They won't try and diagnose the reason why. They'll stop trying and they'll be stunningly hesitant to ever try again, at least with that store and that app. They'll be angry at the store, angry at the app and furious that they didn't get their discount. And they'll tell all their friends on Facebook and Twitter and, perhaps, even those in the physical world.

In other words, if the mobile app campaign succeeds, it's likely to then fail. It's truly the Catch-22 of M-Commerce.

Yet another problem for M-Commerce efforts is the technology's relative immaturity. If a consumer in June 2010 walks into a department store, tries to buy a shirt and the POS crashes, the consumer will be unhappy. He will grumble, but quickly get over it. The consumer is unlikely to stop going to that store, nor will he conclude that POS doesn't work for payments. Why? Because of a generation of generally positive experiences with POS.

M-Commerce has earned neither such loyalty nor such trust yet. That's why mobile experiments are so risky today. One blowup, and consumers are quite likely to abandon mobile and dismiss it as unusable.

Think this is all hypothetical? Not anymore.Think this is all hypothetical? When Apple launched its new iPhone 4 this week, Steve Jobs' demonstration of video chat and browsing the Internet crashed repeatedly. The problem: Hundreds of WiFi devices in the audience were competing for bandwidth with the demo. And last month, Google hit a similar problem when it unveiled Google TV. If the biggest names in technology run out of bandwidth for their wireless application demos, why should any retailer believe that the seemingly bottomless barrel of bandwidth won't run dry for them, too?

Our cynical IT exec reader was especially frustrated at in-store mobile efforts. "Do vendors understand that they need to build partnerships with mobile carriers for microcell deployments in-store to assure the coverage they need to do their server-based offer management? The last time I walked into a Best Buy, by the time I reached the digital cameras I was 'searching' on two devices on two different carriers. So I fire up the Shopkick client, take my barcode picture, feel the rush of anticipation and—nothing. Nothing, that is, until I leave the store, disgusted, when suddenly I regain coverage and 16 coupons and offers stream in. This is, of course, after I have bolted the kids into the car seats. Now that is a positive differentiator. Not trying to be negative, but a failing for years has been a general lack of awareness that this stuff is based on radio—and radio follows the laws of physics."

That's exactly what too many retailers face when they plan applications that require the use of smartphones, WiFi, GPS and other radio-based technologies. Fascination with a smartphone's capacity to instantly deliver coupons won't change the limits of the mobile phone network. And WiFi's popularity is its own worst enemy, if you're depending on wireless networks to support your apps.

Consider a typical shopping-mall retailer. A customer walks in with a smartphone that includes WiFi and GPS. It's the perfect opportunity for an application using GPS to detect that the customer is standing in Men's Apparel and instantly deliver coupons to her phone, right?

Nope. First, GPS is a great way to locate things that are outside: cars, boats, planes and people. But it depends on signals from a network of satellites 12,550 miles up in the sky.

Those signals can't punch through the metal roof of a typical big-box store, never mind the layers of concrete and steel that stand between a shopping-mall customer and the sky. That means that, while GPS can identify a location to within an accuracy of 65 feet outdoors, it's useless inside most retailers.

It's a pity, because GPS is a nearly perfect technology from the point of view of retailers. It's military grade. It can be used by an unlimited number of people at once. It's already paid for and is maintained by the U.S. government. It costs retailers nothing extra and it's already in many of the phones customers have paid for and are carrying with them in stores. Its only drawback: It won't work inside those stores.

Then there's the smartphone. Unlike GPS, mobile phone signal repeaters are often installed in major shopping malls. In the U.S., there are 180 different mobile carriers, according to the wireless association CTIA. But most of them use the bandwidth of Verizon, AT&T, Sprint and T-Mobile.

That means if the mall is big enough and there are mobile repeaters and those repeaters kick out a strong enough signal to reach inside a retailer's store and there aren't too many people in the shopping mall using their smartphones for their own purposes, a smartphone-based coupon application might work.If mobile connectivity is limited, however, that app might not work at all. Or it might work just well enough for a customer to signal her position but not receive the message with the coupons while she's still in the store.

WiFi faces a different set of problems as a basis for mobile applications. More channels are available for WiFi, and WiFi's limited range might allow a retailer to use lots of WiFi access points to blanket a store with a WiFi network. But while taxpayers shell out for GPS and mobile users pay their own phone bills, it's the retailer who pays for that blanket of WiFi: access points, network cable, switches and servers.

Meanwhile, WiFi just keeps getting more popular as an easy way to set up quick, ad hoc networks. That's why Steve Jobs' iPhone demo kept crashing. "There are 570 WiFi base stations operating in this room," Jobs told the crowd after Apple engineers identified the problem. "We can't deal with that."

But WiFi also has security and, much more importantly, security perception issues. "Do you think your customers are ready to acquire, trust and provision your WiFi network on their smartphone device when they enter your store?" asked our cynical reader. "Are you going to offer generic Internet access in your store via WiFi (become a hotspot) or whitelist to only your offers?"

All of these details are part of the IT foodchain of control. Retailers have neither control nor influence over GPS. If it stops working, there's little a retailer can do about it. One small step down are carriers' cellular networks.

If an E-Commerce site is anticipating huge traffic and bandwidth demands, the IT department can do quite a bit to secure more bandwidth and prepare for the deluge. But if an upcoming event is likely to overload the Sprint, AT&T or Verizon networks in one part of town, there's precious little that can be done. If it's a big enough deal and there's plenty of time, very expensive arrangements can—in theory—be made. But it's unlikely.

Retailers can, however, have a huge amount of control over WiFi. But there's still the hurdle of getting consumers comfortable with using your WiFi network.

Ultimately, the best way to deal with the M-Commerce paradox is via mobile moderation. If you're testing a mobile payment method, you do not want huge audiences pouring in at once. Keep your marketing efforts to a low level, perhaps allowing some blog references and pamphlet distribution work their magic.

When you invariably have to go larger, make sure the marketing department uses all of its tricks to spread out the load. E-mail blast offers that are time-limited and stagger the invites. Perhaps do only one department and push only limited items. You don't want 5,000 customers descending on your pilot store from 1 to 2 P.M. You want them spread out over days or even weeks.

Those same marketing tactics can also gently nudge customers into becoming more comfortable with WiFi, by potentially offering guarantees of security coupled with strong financial incentives to use it. Perhaps a 15-percent discount for WiFi purchases, compared with only 10-percent for direct? If you can nudge customers onto WiFi, you have the ability to build more infrastructure yourself—something that's not an option if you depend on mobile networks.

None of this makes wireless retail apps impossible. It just makes it more challenging--as long as you keep your enthusiasm in check. Just don't try ignoring physics.