Target Adds a Ripple to Image-Recognition Pool
Jul 31, 2014 10:56 AM PT
Target last week introduced In a Snap, a new image-recognition app that lets users buy products from the pages of select magazine and printed ads. It is unlike earlier variations of such apps in that consumers don't have to scan codes or follow a link to buy a product.
It works like this: You select an ad and take a picture with your mobile device camera. A snap sound alerts you when information about the product is displayed, and it can then be purchased from Target on the spot.
It is similar to -- although far more limited than -- the Firefly functionality in Amazon's recently released Fire Phone. With Firefly, all a user has to do is aim the camera at a product; Firefly identifies it and then allows the user to buy it on the spot -- from Amazon, of course.
In a Snap works only with specified ad campaigns and media, such as Target's back-to-college catalog -- but it does represent an interesting advance in the use of image recognition, especially for apps that target the everyday shopper's price point.
Not Prime-Time Ready
Image recognition has been languishing in the not-ready-for-prime-time category for some time. Even as it debuts front-and-center on such consumer-friendly vehicles as the Fire Phone and Target's In a Snap app, it's not clear that it's quite there yet.
"It is close, but image recognition still has advancements that need to happen before consumers will use it as readily as they do a mobile camera," James Brehm of James Brehm & Associates told the E-Commerce Times.
In a recent demo, an image-recognition application managed to identify him without his glasses or goatee, and minus 75 pounds, he recalled.
"That is what it should look like," Brehm said. "Fortunately, while we are not there yet, I do believe that ideal is not far off."
In the beginning, though, consumers will have to have the latest in smartphone hardware and software to use the technology as it is intended, he cautioned.
A Look at the Future
Because its hardware was designed with this functionality in mind, Amazon's Fire Phone likely will represent the first large-scale demo of what image recognition truly can do for consumers.
Firefly promises to do away with cumbersome QR and barcode-reading apps for print and real-object identification. It can read email addresses and phone numbers from posters, magazine pages or printed signs. It can recognize songs, movies and TV episodes from just a snippet of audio.
More than likely, brick-and-mortar retailers will find a way to circumvent the worst effects of this technology, speculated Charles King, principal of Pund-IT.
Several years ago, when barcode apps first appeared, retailers like Best Buy would counter the trend by covering bar codes on the products -- or later, directing their manufacturers to use proprietary bar codes that directed a shopper back to the retailer.
It is difficult to imagine this technology being used to help consumers shave off a few dollars from prosaic purchases such as food or cosmetics.
In such cases, "it would have to be really, really seamless and effortless for people to take the time to use it," King told the E-Commerce Times.
For higher-end products, where the savings can be significant, image recognition definitely will have an impact once it is perfected and goes mainstream, he said. "I think in those situations, we will see greater adoption of the technology -- at least in the beginning."