Evan Schuman's StorefrontBacktalk

As mobile computing devices start invading in-store operations?both what customers bring into the stores and what the stores will hand to those customers?the very nature of in-store networking is likely to undergo a dramatic change.

Consider this scenario: A customer walks into a favorite retail outlet and grabs a handheld shopping helper from a rack near the entrance. The customer signs the unit out with a loyalty card and the unit accesses shopping lists from what the customer downloaded onto a Web site and uses in-store directional signals to navigate the customer aisle-to-aisle, displaying context-relavant video ads.

So far, nothing unusual, as this has been discussed extensively. But with the next wave of in-store networking, this scenario takes a science fiction-like turn. The handheld unit can communicate with RFID-enabled smart-shelves and their tagged item-level cousins, as well as signal digital signage to change based on the CRM profile of the approaching customer.

As the handheld device scans pricing, the smartcart acknowledges when an item is dropping into (or below) it. At the end of the shopping trip, the consumer quickly moves to the self-checkout area and barely slows the cart?s velocity. Think of a car approaching an EZPass booth, chuckling at the idea that he needs to slow down to 10 MPH. The smartcart confirms the prices with the handheld device and wirelessly reports this confirmation to the self-checkout system, which greenlights the customer through as it charges the credit card associated with that loyalty card. A receipt is printed as the cart zooms by.

The IT question: Is this scenario a dream or a nightmare? (As a parent, I would add a further choice. If a nightmare, is it the regular kind or one of those night terrors? To put this into context, in the world of retail tech, RFID deployment is a nightmare. PCI compliance is a night terror.)

Are the potential CRM and customer service advantages offered by such an approach going to make up for the mountains of glitches when a store has more than 100 of these wireless devices simultaneously talking back and forth?

There?s no question that the tablet to smartcart to smart shelf to item-level to digital signage to self-checkout reality is still years away, if it ever happens at all. But we?re seeing product rollouts that are starting to go in that direction. Perhaps unexpectedly, the initial rollouts of such devices will likely force the acceleration of other rollouts, as they start to become dependent on each other.

For example, one major vendor is preparing to introduce a series of new customer PDAs that will indeed do pricechecks and other basic functions, but its initial ability to communicate with other devices will be nil. One of the advantages it will be marketed for will be accelerated checkout because the items will already have been scanned.

But that checkout acceleration only happens if the customer uses a staffed lane. If the customer does what the store wants and uses a self-checkout lane, either the attendant has to stop and verify the cart?s contents or the customer must take everything out and lay it on the conveyor belt so the self-checkout system can weigh it, thus defeating much if not all of the convenience.

Was talking the other day with Sheldon Safir, a director of product marketing at Morotola and an unusually insightful person on issues of in-store operations. (This is not to suggest that most Motorola managers are not insightful. They simply do a much better job at hiding any brain stem activity. Safir slipped up and revealed extensive thought operations. He?ll learn to fit it eventually.)

Safir argued that the kinds of customer-issued devices?which Motorola makes, of course?in this scenario can have extensive ROI if retail IT execs are flexible. He envisions the next-generation of devices?which will likely be sold to retailers in bulk at about $700 each?to be used by customers during peak shopping hours and then will be used by store employees (who will switch to a different application) on the storefloor and in the backroom or loading dock for stock lookups and pricing changes. At particularly peak periods, the devices could be borrowed by employees and turned into line-busters before returning it for general consumer use.

Kiosks are another growing area for in-store, but Safir cautioned that retailers have to deploy them cautiously. There is a fuzzy line that separates improved efficiency from decreased customer service and many retail CFOs want to place that line at a very different point than their customers.

There are, however, extremes where the choices are easy. ?A Nordstrom (associate) is not going to tell a customer, ?Go look it up on the kiosk,?? Safir said.

But even an instore kiosk where it is desired can run into problems, as what BestBuy recently discovered when it placed two Web sites in its stores, with one showing the Web and the other looking identical but showing higher in-store prices. The Connecticut Attorney General is investigating whether BestBuy defrauded customers by confusing them as to which site was which.

The other problem with these expanding in-store networks is that they are going to overwhelmingly be wireless. It’s becoming quite clear that wireless networks of all kinds will be an ongoing retail security Achilles heel, as TJX is teaching us all.

What security holes will be opened when consumers are issued devices with limited?and theoretically controlled?access to store information? It’s straightforward to keep unauthorized people out of a network. (Not easy, but straightforward.) But it’s quite difficult to let them in a little but to set up strong barrier against them moving in a little further.

It can be done, but it?s making life a lot more difficult. Common sense precautions can be made, as limiting the wirelessly-broadcast data to a relatively innocuous customer ID number, with the identification, authentication and payment info locked in the server and will then be sent to the credit card processor through a wired connection.

Putting security aside, the allure of networking is that everything seamlessly shares information. Less romantic is the reality that connecting different kinds of devices exponentially increases the chances of a glitch. What’s the favorite network repair headache? That’s when there is a problem with data exchange between Device A and Device B, made by different vendors. Testing shows both devices?on their own?work fine, which gives cover to both vendors to say it’s not their fault. The error occurs when the two devices try to communicate.

I’m not suggesting that elaborate in-store wireless networks of the future are to be avoided. Merely saying that they are far from Nirvana. They are closer to New Jersey. But maybe, just maybe, we’re talking the better parts of New Jersey.