Evan Schuman's StorefrontBacktalk

Amazon is far from alone this holiday season is pushing some new mobile efforts, standing alongside Walmart, Target, Gap and Sears in the popular holiday "let’s fling random things at the cellular tower and see what sticks" game.

But as is Amazon’s tendency, its experiment is a little bolder and more daring. Its not merely that Amazon’s key effort is iPhone-specific—Gap and Target also went for iPhone apps—but that it’s most interesting offering leverages iPhone’s camera and Amazon’s "solution in search of a problem" nominee, Amazon’s Mechanical Turk program.

The new gambit is called Amazon Remembers, which may not be the best name choice for today’s privacy-paranoid CRM-fearful consumer. The idea is for consumers to take pictures on their iPhone and to then transmit the images to Amazon, which will give it to an army of freelancers to try and identify the item and find it—or something really similar—in the Amazon product list.

Although Amazon hasn’t addressed it—and may indeed not be doing it—some are speculating whether the information attached to the photo could ultimately include the exact location that photo was shot, courtesy of the phone’s location finder. Such geolocation data could indicate if a consumer was looking at an item in a particular retail location, which could be very valuable competitive data.

The concept is interesting, but we’re having a difficult time coming up with a viable ROI for it. If this was designed to be an interesting app to give consumers something to play with, it should work. But if it’s intended to actually generate revenue for Amazon above and beyond revenue that they would have likely already received, it’s much more problematic.

In a sense, there are only two broad categories of when a consumer would use such functionality: in a store, home or work environment, where the consumer almost certainly knows what the item is; in a less-familiar environment—such as ballfield, train station, airport or hotel lobby—where the item is more unknown.

The first category is the most likely. A consumer is shopping in a brick-and-mortar and sees an interesting item and wants to price-compare. Or maybe the consumer wants to buy that specific item, but notices that the lines at checkout are so long that maybe E-Commerce would be a simpler way? The consumer pulls out his/her iPhone and the route home is mere minutes away. So far, this is practical and could even be strategic.

The problem is the next step. The consumer already knows precisely what the item is and the item’s box—if not the shelf tag—will reveal precise model and configuration details. As others have observed, an iPhone barcode reader would be far more practical.

With the information the consumer already likely has, the Web browser could quickly go to Amazon’s mobile site and search for the item and grab it right away. For those who are open to the idea of making an Amazon purchase, the time to download and install this Amazon Remembers applet, take a clear photo and to transmit it to Amazon (even if the app automates much of the image transmission) doesn’t seem to make sense. Wouldn’t it take much less time to do a quick Amazon mobile search?

But, a colleague countered, what if the consumer was at a baseball game and saw someone in the stands with this really great-looking watch or perhaps an attractive and warm-looking coat? Wouldn’t it be great to be able to snap a quick picture and have Amazon identify exactly what the item is and send over a link?

Yes, it would be great. But would it be practical? Under the baseball game scenario, that photograph is going to be a far-away low-resolution image. Other than saying "Yep, that was a watch, alright," it’s unlikely if the far-away shot would give nearly enough image details to make an accurate identification possible.

The same goes for someone seen walking down a city street. To get a close-enough image of the item, the iPhone-using consumer would probably need to ask the desired-item-wearing stranger for permission. If they’ve gotten that chummy, it’s probably just as easy to ask, "May I ask what that beautiful watch is?" It’s not materially more intrusive than asking, "Pardon me, but may I take a close-up picture of your watch?"

Please don’t take this the wrong way. I personally applaud Amazon for repeatedly trying new things, new ways to interact with consumers. But before other retailers look at this mobile effort as a nice way to boost revenue, they should question a little more just who would use this and why.