Evan Schuman's StorefrontBacktalk

What if having wireless in-store access isn’t really that important? Retailers’ efforts to make sure customers have constant Wi-Fi access—to fuel mobile functions such as barcode scanning, demo watching and, potentially, even mobile wallet efforts—has certainly proven problematic, whether the reasons are wireless-unfriendly old buildings or young shoppers gulping all of the bandwidth with movies or games.

Beyond encouraging shoppers to use over-the-air access that chains need do nothing to facilitate, what if apps used the mobile device’s memory to play those demos and to look up those barcodes, and then waited to update until the device was reconnected? Shopkick is using one version of this modified store-and-forward mobile strategy, as of an update deployed last month.

The change involved “a massive engineering effort to make this possible,” Shopkick CEO Cyriac Roeding said, according to Payments Source. “A lot of the processing happens now in the client, on the handset.”

Roeding echoed the key Wi-Fi retail struggle (“A lot of [retail stores] have thick walls, where you have a lack of connectivity—but your rewards depend on connectivity”), but Shopkick’s new approach raises some interesting in-store possibilities for all types of mobile apps.

The memory on today’s newer smartphones is substantial (current iPhones max out at 64 gigabytes). Granted, no one can know how much of that is available, given other apps and media stored on a shopper’s phone. But there’s ample room for SKU databases, some demos and other items that will be needed for in-store offline action. The nice part is that data can be put on standby and can synch up with headquarters whenever a network connection is reestablished.

First, in-store mobile performance will likely accelerate and could accelerate sharply. Second, the struggles of splitting store Wi-Fi bandwidth (reserving a healthy chunk for associates’ tablets and internal needs) and making sure there’s enough for shoppers can be radically simplified. (The IT motto: If you can’t solve it, sidestep it.)

Third, this strategy will enable much more universal mobile rollouts, given the lack of a need to rely on each store making Wi-Fi modifications.

In essence, this approach—and Shopkick deserves a lot of credit for making this non-traditional move—requires a sharp rethinking of mobile strategy. Although many chains have thrown out lots of ideas for truly real-time interactions using mobile, few have found the tech hurdles involved to be worth the effort.

Walmart SVP for Global E-Commerce Venky Harinarayan, for example, spoke of Walmarts with big screens where associates and customers chat back and forth about products.

Shopkick’s application works through the phones listening for an audio signal broadcast in the stores by a basic speaker system. The frequency is such that humans can’t hear it, but the transmission doesn’t involve wireless. It awards points for the customer being in the store, alerts the potential shopper to special offers, and then awards points for various actions, such as scanning barcodes and trying other activities. But there’s no particular need for those points to be sent to the central server right away. A delay of minutes or hours—or even, theoretically, days or weeks—shouldn’t make much of a difference. There was the epiphany.

So many of today’s mobile interactions do not really need to be real time. Consider QR codes. With the exception of some theoretical not-yet-deployed tactics—such as an approach that was briefly considered by Home Depot and Macy’s to have purchase histories impact what a QR code would display—QR codes have a limited number of responses. Why not store them in the app? And then use the later connection to update central databases?

RFID applications wouldn’t use wireless anyway—we have seen hardly any RFID apps used in consumer mobile apps anyway—so that wouldn’t be an issue.

How about NFC payments? The POS device would have to communicate in real time for payment approval. But that communication would use the internal LAN Wi-Fi, which is not the problem. The more common special offers could be updated to the phone for offline display. But the easiest route is to simply wait until the customer’s device is reconnected, and then display the offer.

This shouldn’t be an issue. A post-purchase special offer or message isn’t likely to impact a current shopping trip anyway, so a short delay won’t reduce its effectiveness.

In-aisle purchases, which are just starting to happen, could mostly be done using broadcast cellular signals. That would work for many chains, but not all. Macy’s flagship location at Times Square in New York City, for example, is notorious for blocking just about all signals from outside the building, but it’s fortunately the rare exception.

Although geolocation/geofence capabilities need a constant connection, they rarely work in-store, so it’s not a problem. The handful of in-store efforts to locate customers has generally struggled to be useful, given the lack of precision. Finding an aisle is not that difficult. Finding a precise spot in that aisle is much more useful.

How effective have initial efforts been? Not very.

A Meijer trial used three location techniques (a planogram, Wi-Fi signal strength triangulation and product-barcode scanning) layered on top of each other and was still not getting better than 30 feet of precision. A recent Google effort didn’t perform much better, with dozens of feet of inaccuracy. And in December 2011, a Google Maps trial with Home Depot to zero in on product location glitched big time when it started sending people a few blocks away—to Home Depot arch-enemy Lowe’s.

The point is that temporarily setting aside Wi-Fi-based customer location efforts may not be such a loss. When the technology improves to the point where precision gets down to the 1- to 2-inch neighborhood—where item-level tracking could zero in on the specific flavor and brand-name of cereal desired—things can be rethought again. Mobile strategy needs to be a constant compromise between functionality desired and functionality that can be reliably delivered. For now, Shopkick’s store-and-forward approach seems to be very much worth immediate consideration.