Evan Schuman's StorefrontBacktalk

Another reason: Sites are inevitably easier to upgrade than apps. A Web site fix or improvement just has to be developed and tested, and then go live. An app, on the other hand, has to be developed, tested, and then sent to the app store, where it may be vetted by the store’s owner, could be rejected and will go live on an unpredictable schedule. Then it has to be downloaded by customers, who may or may not get around to it.

And all that extra work has to be done for each version of the app. No wonder BlackBerry and Windows Mobile users are stuck with the Web, and apps that merely mirror what a Web site does are always a generation behind the current site.

But for retailers, an app naturally does at least one thing better than the Web: It blocks out the competition. The mobile Web makes it easy for customers to jump around and visit the competition—it’s like being in a mall on roller skates. A retailer’s app, on the other hand, puts a customer firmly in a single store, and the retailer can control the customer experience almost completely. That may explain why Nielsen found app users spent more time at each store. In an app, there are fewer distractions.

That’s probably not enough differentiation to make apps worth the trouble for most users, though. An app really needs to do something that makes it worth the trouble. Pizza chains have figured that out—they’ve made a priority of putting ordering capability into apps. You can order a pizza through the mobile site, but the app makes it much easier.

But purchasing isn’t the only function apps might deliver on better than mobile Web sites. With ever-increasing processor power in smartphones, a walk-through virtual store becomes practical. That’s never going to be as smooth and effective on a mobile site as it is within an app that is completely under the retailer’s control.

Or maybe the app could leverage its ability to tap into the phone’s location, motion sensors and camera—things a Web site can’t count on being there—to provide in-store-like makeovers or, in the case of chains with lots of locations, make use of geofencing for coupons that are good for very limited periods. (Related story: Gap’s Geofencing Trial Merely The Appetizer Before The Purchase History Entrée)

Of course, if there’s really nothing special your app can do, it’s probably not worth the trouble. Remember, only about 8 percent of those Nielsen smartphone users were app-only users. The rest used the mobile Web at least some of the time.

But pulling customers in and blocking out the competition is what apps do best. And as long as you can find something compelling that the app can do, it’s likely to be worth the effort.