We’ve written articles highlighting shoppable images as well as image recognition for fashion, but never have we compared the emerging technologies side-by-side. And in all honesty, that’s not exactly what we’re going to do here, as it might not be fair to pit them against each other since they’re so different in nature. But, we can certainly examine the benefits and drawbacks of each, which will help us to gauge the usefulness of each technology.
While more specific than “shoppable content,” the term “shoppable images” could still mean many different things to different people. A number of companies have their own take on the idea (even Echidna is developing some exciting new shoppable image technology), and only time will tell which model drives the most results and actually ends up becoming prevalent in the market.
Although the concept of shoppable images is fairly fluid at this point, there are several benefits that stand to be gained through its use. Perhaps most substantially, shoppable images provide inspiration for online shoppers by showing the product in context (as opposed to a standard, lifeless product image). Additionally, shoppable images allow multiple related products to be shown concurrently, which could increase the number of different items being added to the cart. Shoppable images can also be integrated straight into the retailer’s website, thereby reducing friction and shortening the path to purchase, which should result in higher conversion rates than image recognition (we’ll touch on that more below). The main drawback to shoppable images lies in the fact that the shoppers’ imagination is inherently going to be limited by what’s on the screen. For shoppers more accustomed to the traditional method of browsing through product listings, shoppable images could also be confusing and/or frustrating.
VERDICT: The concept of shoppable images looks to be very promising if it continues to be developed and is implemented effectively. We’re definitely excited about the possibilities surrounding it.
If you look at image recognition to find exactly the same items, you will be disappointed.
First of all, people are probably always going to be disappointed with image recognition, at least in the foreseeable future. Many have tried to tout image recognition as a magic solution, when in reality the technology just isn’t quite that magical (at least, not yet). This idea is summed up nicely by Daniela Cecilio, the CEO/founder of a fashion recognition company called ASAP54, when she said “If you look at image recognition to find exactly the same items, you will be disappointed.” Instead, image recognition as it currently stands is meant to merely show similar or related products.
Like shoppable images, image recognition gives inspiration for shoppers, which is its strongest attribute. And instead of simply using images on a screen, image recognition meets people where they’re at, merging the real world of the shopper with the retailer’s digital presence. As hinted at above, though, some image recognition systems have definitely had issues with checkout complexity: while shoppable images are integrated directly into a retailer’s site, many image recognition platforms have had referral elements, adding extra steps into the purchase process, ultimately leading to lower conversion rates. The full results are still to be seen, but retailer-specific image recognition apps (such as Zalando’s) could circumvent the referral-related difficulties. Another main drawback to image recognition has already been mentioned, but essentially it’s the misunderstanding of the technology and what it can/should do, as a lot of people incorrectly expect it to be a visual search engine that spits out exact product matches.
VERDICT: Image recognition is definitely another exciting retail technology, but like shoppable images, the concept hasn’t been fully realized yet. Expect the technology to continue to improve, though, and effectiveness with it.