Is image recognition technology poised to revolutionize omnichannel fashion retailing?

In the age of omnichannel retailing, the rapid adoption of new technologies has made it possible for retailers to deliver a consistent message and experience to every consumer, no matter how they choose to interact with a brand.  Consumers are no longer satisfied simply visiting a brick-and-mortar location and completing a purchase.  They want to follow companies on Twitter, check reviews of products online before going to the store (reportedly 80% of store shoppers checked prices online in 2013), comparison shop on their smartphones as they browse, and complete transactions with their iPad.  In this omnichannel environment, many retailers have appropriately responded by creating entire teams to manage social media accounts, build e-commerce sites, and optimize supply chains to serve consumers online and in-store.

Over the last few years, more and more retailers have been embracing a new technology to further deliver on these consumer expectations – image recognition and visual search.  Recognizing that consumers often want to look up product information while in-store or on the street, marketers began by leveraging barcode scanners and QR code readers to direct consumers to PDPs, sometimes even allowing them to purchase online.  For various reasons that would require a separate article to fully explain (QR app adoption rates, inconvenience of finding and scanning codes, lack of consumer education, etc.), these interventions rarely aligned with consumers’ desires and often fell short of forecasted conversion rates.  Instead, retailers have recently begun utilizing image recognition technology to skip the QR code and allow consumers to leverage a technology nearly everyone already has and is comfortable using – the cameras on their smartphones – to take pictures of products they are interested in and to immediately be directed to product detail pages.

Amazon was one of the first major retailers to leverage this technology for product discovery at scale.  First with its Flow app and now with the Firefly feature native to the Fire smartphone, Amazon has been at the forefront, pushing consumers to use this new technology to recognize products and purchase them within the company’s owned marketplace.  However, Amazon’s Firefly feature to date has been primarily focused on text recognition for its visual searches (it also has Shazam-like features to identify music and videos), and this limits the type of products the app can identify to products with clear labeling and packaging, such as video game covers and soup cans.  Given my years of experience in the footwear and apparel industry, I only became actively interested in the power of visual search and image recognition after seeing the technology applied to fashion.  Leveraging technology developed at universities such as the Imperial College London, early entrants into the space created applications like Snap Fashion, Style-Eyes, and Slice, which allowed users to snap a photo of a sweater or dress they liked and then receive recommendations from affiliated brands within the app’s network for similar products based on color, shape, and pattern.

Rather than search across an entire marketplace of disparate brands (as is the case with Amazon and early app’s like Snap Fashion), there has been a recent trend for individual brands or retailers (Adidas, Target, and Macy’s to name a few) to launch applications leveraging this technology to lead consumers to products within their own databases and enable mobile purchases.  Given these retailers’ desire to satisfy their consumers’ omnichannel needs, this investment makes perfect sense.  Leveraging third-party technology from visual search companies like Cortexica (among others), retailers can add image recognition features to their already existing mobile applications.  This has numerous direct benefits for the omnichannel consumer.  Image search features enable product discovery at the moment of inspiration, wherever a consumer has the chance to snap a photo.  This technology can be linked with the backend of e-commerce sites to enable immediate conversion and impulse purchases.  It can also provide valuable product information for consumers while they are in a retailer’s brick-and-mortar location, further blurring the lines between the physical and digital consumer experience.  Individual retailers gain an advantage by creating their own, proprietary database of searchable images, ensuring consumers are given recommendations for only their brand’s products, perhaps even pushing specific content depending on inventory and margin information.

We are still in the early stage of retailer adoption for this technology, and big questions remain around whether or not consumers will adjust their search and discovery habits to incorporate these types of features and whether an aggregator or a large set of customized app’s will dominate this space.  However, it is clear this technology has the potential to have a huge impact within the fashion world, and I am excited to see how this environment evolves.

Sources

http://marketingland.com/why-brands-should-go-omni-channel-in-2014-70970

http://www.mobilecommercedaily.com/macys-significant-omnichannel-push-includes-showroom-busting-image-search-app

http://techcrunch.com/2014/06/18/amazons-fire-phone-introduces-firefly-a-feature-that-lets-you-identify-and-buy-things-you-see-the-real-world/

http://www.mobilecommercedaily.com/target-fires-back-at-amazon-with-its-own-image-recognition-app

http://hypebeast.com/2011/8/adidas-originals-launches-iphone-app

http://www.snapfashion.co.uk/

http://www.cortexica.com/technology/