With costs rising and margins falling, retailers will increasingly need AI to help them improve the customer experience, automate interactions and improve their knowledge of their customers
Competition in the retail sector is getting tougher. Physical stores are having to respond to online retailers becoming direct competitors while online retailers are having to get used to selling in a crowded market. To help them in their fight to keep revenues high and costs low, they’re now looking to technology to help them become more effective. This often now means using artificial intelligence and machine learning, and this article describes some examples of emerging uses for them in retail.
For many years, customers have dreamt of saying “I want some shoes that look like this” – and now they can. Visual search companies, such as Slyce and iLenze, are providing the AI based technology to turn an image from a shopper’s smartphone camera into a nearest match from the store’s product catalogue.
The concept uses neural networks which are trained by identifying the distinctive components, known as classes, in the images of a store’s product catalogue, such as their shape, colour and proportions. Then, when a shopper uploads a photo of something they want, it gets broken down into those classes and a catalogue search find the closest match.
That process isn’t too dissimilar to how a human would do it, but an AI based service can do something online that would have previously needed to be done in store or do something in store using a cheap machine that would have previously needed an expensive human. Customers, physical stores and online retailers are all winners.
Although we regularly hear about personal digital assistants appearing in our personal lives - such as Siri, Cortana and Amazon’s Echo – they’re also appearing in stores. These intelligent assistants, known more technically as AI based chatbots, provide Q&A based interactions such as helping customers query real time stock levels or asking for more information, such as “does what I’m looking at also come in blue?”.
Using AI to integrate natural language processing, which converts a conversational question asked by a human into a technical query a database can understand, may seem like a quick win for online retailers but how will physical stores benefit? Some are installing physical machines in their stores but perhaps the surprising answer is that they’re increasingly starting to make their physical stores an extension of their online stores. Customers, especially the next generation, increasingly want to walk into a store and take out their smartphone. Already Uniqlo allow customers to walk into its most modern stores and ask Siri questions about the stock that’s in front of them. Impersonal for the customer? Probably. Cost effective for the retailer? Definitely - so expect more of this in the future.
The goal of most retailers is to provide a bespoke service at a commodity price point. Knowing what a customer probably wants to buy and when doesn’t just help the stock department but also makes the customer feel recognised, welcomed and wanted. Online retailers have a head start over their physical peers in collecting the data needed to make this happen as they log everything that gets viewed never mind bought, but loyalty cards are helping physical brands catch up.
Once sat on a mountain of data about everything a customer has looked at, purchased, the weather, the news, social trends and much more, a retailer can introduce machine learning. Unlike broader AI technologies which need to be told what to do or what models to use, machine learning uses training data to work those out for itself. By finding the models, machine learning helps predict unknown values when other values are known. In this case, who might want to buy what and when. This level of deep insight recognises that what someone wants to buy while browsing on their way to work might be very different to when they’re at home on a laptop. Same person, same buying history, but very different buying intentions.
Using machines to scale out human interaction
The reality then is that the retail experience is becoming a digital retail experience wherever it happens and proving it isn’t just a job for just humans any more. The power of machines – by using AI – is being used to provide a mobile experience as well as a fast and timely experience. But most importantly, for retailers it’ll soon be needed to provide a competitive experience.