How Artificial Intelligence Can Unlock the Potential of Mobile Retail

Reposted from Internet Retailer.

By Andy Narayanan, Vice President of Intelligent Commerce at Sentient Technologies

AI is uniquely able to provide a text-free shopping journey, one that transforms mobile shopping into a wordless visual conversation.

As much time as we spend on our cellphones, we actually don’t use them to buy all that much. According to eMarketer, while mobile phones boast about 30% of all retail traffic, they only account for 11% of revenue. We browse on our phones, sure, but when it comes to buying, we actually use our desktops or head to a brick-and-mortar location that carries products we’re interested in.

So why is that exactly? For starters, it’s simply easier to navigate retail sites on a laptop or desktop. Search is still primarily text-based and, since most of product searches require several iterations (choosing color or brand or just outright searching again when our first attempt misses the mark), a physical keyboard is just easier.

A subtler and more interesting reason is that mobile retail sites haven’t actively leveraged what should be their advantage: the device itself. Nowadays, for most people, our phones are on us constantly. We pull them out to settle little disagreements (was Kevin Bacon in that movie?), check the news while we’re waiting in line, get movie tickets so we don’t have to wait when we get to the theater, and on and on. Most mobile sites don’t take advantage of this. In fact, most are just responsive versions of a retailer’s browser site.

So if mobile doesn’t convert at the rate it should, how exactly can we get it there? How can we turn mobile searchers into buyers? How can we alleviate the shortcomings of mobile while maximizing its inherent advantages? The answer lies in a field that’s been exploding over the past few years: artificial intelligence.

Let’s start with the shortcoming we mentioned: interface. Searching on mobile simply takes us less time than shopping on an actual computer. There’s less screen real estate and no keyboard. Products are just harder to find. But AI is a good solution for this problem. For starters, AI is uniquely able to provide a text-free shopping journey, one that transforms mobile shopping from a clunky, faux-browser experience to a wordless visual conversation. And when you think of the experiences that really sing on mobile, you likely think of social apps like Instagram, endless streams of clickable images and personally relevant information. That’s what AI does incredibly well. Here’s how it can work:

Instead of going through the typical search experience, all you’d need to do is tap on images of products you like. The AI looks at the images themselves and surfaces similar options based on image alone. Smart AIs understand product facets we typically search for, like color and brand, as well as subtler, more difficult-to-describe characteristics, like fringe, unique patterns and logo placement. Instead of re-searching or clicking tiny text boxes, you’d just tap the next image you’re interested in. Each tap successively trains the AI to understand what you want, right then, whether it’s red shoes or a plaid hat.

But that’s really just the beginning. Because our phones are on us constantly, artificial intelligence is going to be able to contextualize our behavior based on a whole host of important signals. And for it to really fulfill the promise of AI, it’s going to be have to feel completely seamless.

For starters, just think about the mobile shopping experience. It’s a lot different from shopping on a laptop or desktop. As opposed to an end-to-end experience, mobile is more like a set of micro-transactions in a holistic shopping journey. We rarely fire up the phone, find what we want, then purchase it outright. Because we use our phones differently, the shopping experience is simply, well, different. We might find a product on the way to work and save it for later. We might pull out our phone on a quick break to check details. We might show off what we’re thinking about buying to a friend over a drink after work.

These actions are very different from each other. One is discovery, one is refinement, one is comparison. AI is uniquely capable of understanding that context and, importantly, responding with the details and experience the shopper needs right then.

For just a single example, let’s focus on refinement. At this point, a shopper has looked through a few products she likes and now wants some extra details on the product in question. Smart AI agents can pick up on all the contextual clues and offer her the details important to her, right then. Maybe she constantly chooses shoes that are at a discount: in that case, the AI can make a point of surfacing bargains, free shipping offers, or similar shoes at an even greater discount. But what if she has an incredibly up-to-date style and doesn’t seem too bothered by price? AI can help there, too, showing her the newest inventory from the hippest designers, or offering her a membership for exclusives on new products. Every tap and interaction is a signal that an AI can learn from and, more importantly, leverage so that shoppers can find products they love.

It’s important to remember that this won’t feel like AI to users. Much like how Google moved to AI-powered search algorithms without fundamentally changing user experience, AI in retail doesn’t need to take the form of personal shopping assistants or cutesy chatbots. Instead, it needs to react to user behavior and preferences at every level, learning from what they click, browse, and like, and then showing them products that fit their unique preferences in that unique moment of time.

In other words, the future of retail is full, true personalization. Right now, we’re in a world where users are bucketed into cohorts–users like you bought items like this–as opposed to one where our technology actually understands us individually. True personalization knows your patterns but also understands that your likes and dislikes change over time.

For example, AI can understand that when a sneaker shopper is looking at dress shoes, it means something. It might mean a big job interview or a wedding, and it can react to that data in the moment and make the purchasing process easier and better. Maybe it can suggest an outfit to go with the dress shoes. Maybe it can match other signals (like a calendar notification for an upcoming flight) and realize you need the shoes shipped. Or that there are brick-and-mortar locations nearby that match the site or brand that customer is looking at. The list is almost endless. And a lot of it can be enhanced–not hamstrung–by our mobile use.

Which is all to say that AI is coming to retail. In fact, it’s already there. Retailers who integrate AI into their mobile sites, who devote real thought to leveraging the signals sent every day, those are the sites that are going to stand out and delight their customers. And all that forms a virtuous circle, where a mobile device and a retail ecosystem actually understand individual users. In a way, it will be a lot like shopping at a favorite store with a particular employee who knows exactly what you want. And those experiences are far, far better than wandering the endless aisle.