Several vendors, well aware of many mobile device owners' love of taking digital photos of anything and everything (including selfies, which to me have always suffered from a major lack of raison d'être), actively encourage these shoots, hoping to pull in a goldmine of data. The pitch to shoppers is simple: if you see anything you'd like to buy, take a picture of it and we'll quickly identify it, through software and crowdsourcing.
Whether or not those identifications will work or not—and whether there are much easier and more accurate ways for those products to be identified—is something I'll get to shortly. But the goal here is all data. First, the images are being shipped through a mobile app, so everything is being associated with a specific identified shopper. (Hello, CRM database.) Secondly, the images usually come with exact geolocation data ("Seems that you took this picture in the housewares section of our direct rival on Elm Street. Good to know.") plus date/time.
Much more importantly, the images reveal much more than that handbag you wanted them to identify—beyond the obvious detail that you're someone who wants to spend money on a handbag right now. The vendor knows what was sitting five feet away. Put more bluntly, it knows what the shopper chose to not photograph, which could be even more revealing.
Then there's price. Major retailers have pretty much mastered online pricing, with systems able to instantly detect whenever a rival changes any online price and to react automatically by increasing or decreasing its own price for that product accordingly. But the prices retailers charge in their physical stores, that's still a mystery only broken by either sending spies into the store or by crowdsourcing the prices via cooperative consumers, which is as inaccurate and incomplete as it is cumbersome.
Then we have these thousands (millions? Tens of millions?) of photographs coming, seeking product identification. They often will show pricetags. This may seem trivial, but that little piece of extra data might ultimately prove the most valuable, if it helps stores intelligently adjust their in-store prices. Electronic shelf tags offer the potential for this to eventually be automated, which is lightyears beyond where even the most sophisticated largest chains are today.
With all of these advantages, what's the catch? A big one. There is a very distinct possibility that the incentive being waived in front of shoppers—that this tactic will quickly and accurately identify products for them, for free—simply won't pan out, at least not in many of the attempts.
Why? There are two broad places where these photos could be taken: Inside a physical store; and outside a physical store (on a street, at a friend's house, in a non-retail business, etc.). The "inside a store" option is the one that vendors most want to access, as it promises the most data goodies. But for the shopper, it's likely to make the least sense. If you're standing inside a Walmart, Target, Footlocker or a Macy's and you see a product on the shelf, how difficult is it to identify what it is? There's a box with a description, a label nearby, barcodes to scan (for instant and precise identification) and generally associates nearby. Seems that shooting a picture of a product inside a store to identify it is probably the most labor-intensive, slowest and least reliable option.
Let's move to the street. Now that is where there is the most potential for customer benefit, but it's also the most challenging environment to get an accurate result. Scenario: A woman is walking down Madison Ave. in New York City and she sees another woman with the exact purse that she has been looking for—unsuccessfully—for months. She quickly snaps a photo, the vendor identifies the exact model and where it's sold and all is wonderful. Except that it's highly unlikely to work that way. A more likely scenario is that the purse is spied. The first woman scrambles to grab her phone, open the camera app and frantically tries to refind the original woman with the purse and shoot. The result will likely be blurry and include dozens of people with far more accessories. If she just clicks and sends it, the software will have no idea what to focus on. If the app permits, the shopper has to write a description of the targeted woman and the handbag. This is already getting too labor-intensive.
The purse-carrying woman could be approached and asked if she would pose with the purse, for a clear and focused image. First, done incorrectly and law enforcement will get involved. Secondly, if the purse-carrying woman is that cooperative, why not just ask her where she got the purse? Instant and accurate information will likely follow. (Cooperative person walking down Madison Ave.? OK, that's another problem with this scenario.)
Similar issues crop up with the other scenarios. In a friend's house and you see something you'd like? Why not just ask?
That said, there are situations where it could be helpful, such as visiting an estate or garage sale and you see an antique and you have no idea what it is supposed to do—and the people running the sale also do not know. That could be useful, but it's not going to happen very frequently. Much more importantly, the business model falls apart there. The companies paying these vendors are consumer goods manufacturers (P&G, Nabisco, Nike, etc.) and major retailers. They have no business reason to fund identifying 100-year-old antiques that they don't sell or manufacture.
All in all, this actually has great potential because most consumers will love the idea of taking photos and won't stop to think how unlikely it is that they'll usually get information that they couldn't get easier other ways. But while they're trying, a ton of information is being delivered. The intersection of smartphones and dumb consumers is a manufacturer and retail wish come true.