With Mobile In-Store Apps, Will Reliable Beat Sexy?

In the latest round of in-store mobile app vendor battles, the goal seems to be to capture the title of easiest to use. That's an ideal goal, but "easy" is a word that in tech circles has a deliciously paradoxical nature. The easier and more intuitive an app is, the more sophisticated and complex is its programming. That generally means there's a lot more that can—and will—go wrong and glitch.

From that perspective, "easiest" would be the app with the most basic functionality but that works almost always. So in the payment world, a magstripe is easier to use than NFC, even though NFC is faster and more flexible when it works perfectly.

A company called QThru, for example, is making waves with a mobile app that handles product identification and completes mobile payment through scanning an old-fashioned barcode. Its claim to fame: that the app identifies the product within two seconds even when the hand holding the phone (which is taking the picture of the barcode) and the hand holding the product with its barcode are both shaking and when the barcode is incorrectly positioned.

"You can hold the barcode upside down, sideways, at an angle and it will still pick it up," said Phil Stafford, the vendor's Chief Marketing Officer. "We wanted to make sure that that part of the experience was effortless."

Demos of the product—and a customer who has been beta-testing it—support the vendor claim.

Contrast those statements with ones from IBM and Amazon, where both companies rolled out mobile augmented-reality apps. Those apps superimpose data on the screen above/next to the live image of the product. It could integrate that shopper's CRM data, enabling even more customized information.

The data appears on the screen of the mobile device, as opposed to Google Glasses, where the data would be superimposed on the product itself (or so it would appear to the wearer). (Quick aside: Wonder if Google Glasses will be available in prescription form for shoppers who wear corrective lenses? If not, will it be the ultimate Geek look to wear Google Glasses on top of traditional glasses? Would that be the optical equivalent of a double pocket-protector?)

Let's get back to IBM. IBM's statement was chock full of tantalizing marketing come-ons, such as the ability to overlay same-day product discounts and social media comments from friends about that product. And the ability to quickly locate products with specific criteria, such as scanning an aisle full of products and having the app flag foods that are low-carb or electronic devices that have a battery life more than a predefined threshold.

But Big Blue's statement glossed over how that identification would happen, merely saying "When (shoppers) point their device's video camera at merchandise, the app will instantly recognize products." Instantly? Really? How will that identification happen? Will it truly be able to—consistently and accurately—scan a whole aisle of products, identify those products that meet predetermined criteria, do lookups and superimpose the data—all before that tube of toothpaste has moved out of the screen view?

In talking with reporters, IBMers have vaguely said that shapes and colors will identify the products. What is the app looking for? The product label? The container? What happens when the label or container changes? How close does the shot have to be? How much of the product needs to be in the screen view and for how long? What do retailers have to do when any product needs to be added into the database or changed?

Getting back to the definition of easy, this was the most intriguing comment from IBM. John Kennedy, an IBM corporate marketing VP, was quoted in AdAge saying that he "believes the augmented-reality app may be better received by shoppers than QR Codes or barcode scanning because it's relatively effortless, involving simply holding up a device while walking the aisles or look at products, similar to 'something we do every day. The information is embedded on the product itself, so it's more like a natural behavior,' he said."Effortless/easy only happens when the product works perfectly. Big Blue deserves applause for thinking big and for trying to craft something that could truly advance shopping efforts. But the level of complexity it's introducing is likely to create an initial experience quite far from effortless.

Amazon's augmented-reality mobile app (called Flow) lies somewhere in between IBM's and QThru's efforts. It also superimposes data over products, but its product identification uses reliable barcode scanning and QR code scanning.

QThru is actually part of a more comprehensive service, with mobile payment. The idea is that shoppers download the app, arrange for payment, and then go shopping. After items are scanned, the phone or tablet interacts with a kiosk (there are alternative methods, as well), where a receipt is printed. An associate then matches the receipt against the products in the cart and customers are sent on their way, bypassing POS lanes entirely—in theory. It could also use a self-checkout lane, except that all of the items have already been scanned and the payment has already been processed.

QThru's Stafford said his company was initially using eBay's RedLaser mobile app but stopped because of concerns. First, he said, "it had issues with low light" that caused problems identifying products. Second, he was nervous about what direction eBay may ultimately take for RedLaser. "It was unclear what the future was going to hold for that. Will they keep it public or lock it down and license it, limiting the accessibility?"

Stafford also touched on weaknesses that QThru's homegrown app still has to overcome. "We can certainly scan items at an ultra-quick rate, but there are asterisks. If an item is flat with a prominent and clean barcode on the packaging, it tends to scan much more quickly than, let's say, a wrinkled bag of potato chips," Stafford said. "That is why I chose to show off the scanning on the small can. Anything with glare or hard curves can slow it down. But it still performs as good as or better than RedLaser."

One beta tester for QThru is Tyler Myers, President of The Myers Group, which owns a few grocery stores, three hardware stores and three convenience stores/gas stations. Myers said they installed scales for him in produce, where receipts are printed after fruits and vegetables are weighed and the app scans "some sort of a code that the QThru app reads."

He's also limiting purchases on the system to 10 or fewer items "until we get our arms wrapped around it." But Myers' key question is whether such apps will ultimately increase—or decrease—basket sizes.

There's an interesting psychological rationale for why it might increase purchases, especially after the novelty wears off. Today, shoppers are hesitant to add too many additional items to their shopping cart, because they know of the multiple delays every extra item will cause. That's one more item that needs to be taken out of the cart and put onto the conveyor belt, one more item that needs to be taken off the belt and bagged. By bypassing traditional POS lanes, there is less of a time-crunch penalty for adding items.

Whether customers will realize that, though, is a very different question.

The ease-of-use issue, however, is likely to be the most critical factor in the success of any of these approaches. A customer who runs into just a few glitches during an early test is much less likely to try again. It may not be nearly as sexy as what IBM and Amazon are pitching, but if QThru's approach identifies products more accurately, reliably and quickly, that may just be sexier than superimposed on-screen discounts.