Join gaming leaders online at GamesBeat Summit Next this upcoming November 9-10. Learn more about what comes next.
Artificial intelligence and shopping — does that mean robots that’ll stock the shelves?
Yes, absolutely. We’ll have robots that will use A.I. to check inventory, help customers find the items they’re shopping for, ferry supplies from one part of the warehouse to another, aid with shipping, you name it. But the real revolution for A.I. and shopping will be invisible because the technology will create better experiences for consumers while helping employees and shopkeepers run operations more effectively.
Let’s take any big department store that sells thousands of home, fashion, and beauty products — more specifically, let’s say that there are exactly 100,000 products. Because customers will buy up to 150 percent more and be happier with their purchases if they’re shown the items in context, merchants will create outfits, design window displays, and produce splashy catalogs and digital lookbooks to help customers imagine how to wear the latest fashion trend, how to arrange their living rooms to show off their new velvet sectional, or how to install an outdoor shower.
The manager of that department store wants to highlight all of the store’s products, but only 5 percent of retailers’ inventory typically make it onto a mannequin, into a shop window, or onto the pages of a catalog. And here’s why: Let’s say there are five items in any given product set, and let’s say that there’s no overlap. That means the store would have to create 20,000 sets. If it takes five minutes to make each set, that means the retailer has to spend 1,667 hours to make enough sets for all their products.
The math doesn’t stop there. Let’s say every day .05 percent of the inventory goes out of stock, and every day the store also adds 500 new products, which means they need to have outfits designed around them, digital lookbooks of furniture sets reorganized, or make-up bundles rearranged. That’s about 17 hours each day maintaining, fixing, or building creations anew — forever.
And because it takes the manager and her employees so long to set it all up in the first place (about 70 days, in fact), they have to go back to repair anything that may have been damaged during that time. Mathematically, that means 70 days x 17 hours daily = 1,190 extra hours fixing and adding outfits, cosmetic kits, and so on.
It’s all so painfully, exhaustively manual.
Enter machine learning, the subset of A.I. that teaches a computer system how to learn. So, for instance, at my company Findmine, based at NYU Tandon’s Data Future Labs, we teach our system how to learn why an outfit or combination of furniture works, so it can make its own decisions. It allows us to take our customers’ entire product catalogue and build outfits on the fly or keep up with changing inventory — even changing seasons and trends! — because it’s figured out what works and what doesn’t, constantly readjusting to information.
Machine learning also makes possible many other efficiencies, some of which have already changed the way we shop. For example, it facilitates conversations between shopper and the retailer to automatically answer your customer service question. Or take a company like Truefit: Its technology can understand which brands will best fit a shopper’s body type by observing how shoppers with certain body types buy (and return) particular brands. Higher education institutions recognize the potential of artificial intelligence. Tandon’s Future Labs house a variety of companies in their three incubators, and NYU Tandon launched the A.I. NexusLab along with ff Venture Capital in July to support a group of similar emerging A.I. companies.
Marketing technologies can also predict the optimal time, place, and content for a consumer to receive a message about a product he’s been eyeing. Some retail technologies will also use machine learning to prevent fraud (Sift Science), catch glitches before Black Friday shopping hordes can crash a site (Prelert), or track competitors and know which products to purchase (StyleSage).
All of this is possible because machine learning improves with more data, and retail has lots and lots of that — zillions of products, myriad transactions, and gaggles of customers, each with their own trove of associated data points. That primes the retail sector for disruptive innovation. As shoppers demand more flexibility and better experiences, machine learning and other A.I.-powered technologies will replace the underlying backbone for much of the industry’s activities, resulting in more effective operations. Those cost savings will translate into wallet-friendly pricing, better inventory assortments, more exciting products, and fewer out-of-stock, shipping, or customer-service hiccups.
The shopping revolution won’t be televised; it’ll be quiet and unseen, except maybe for that pesky robot zipping around you at Target as it hurries to locate that gift item it already anticipated you wanted.
VentureBeatVentureBeat's mission is to be a digital town square for technical decision-makers to gain knowledge about transformative technology and transact. Our site delivers essential information on data technologies and strategies to guide you as you lead your organizations. We invite you to become a member of our community, to access:
- up-to-date information on the subjects of interest to you
- our newsletters
- gated thought-leader content and discounted access to our prized events, such as Transform 2021: Learn More
- networking features, and more