If you’ve attended retail conferences over the past few years, you’ve undoubtedly walked away knowing ‘omnichannel’ was on just about every retailer’s priority list. The term quickly became one of the most overused buzzword in retail marketing. This year, we’re hearing a lot of buzz around another term in retail: artificial intelligence (AI).
Retailers such as 1-800-Flowers, Macy’s, Staples and Walmart have publicly discussed their commitments to AI. Home improvement retailer Lowes plans to have roaming robots – LoweBots – in the aisles of 11 Bay Area stores this year to respond to shoppers’ questions, such as “Where are the screwdrivers?” And Starbucks’ new AI-powered My Starbucks Barista feature allows mobile users to place their order by speaking to a virtual barista within the app. The barista bot pings the order to a nearby store where an employee makes the drink for easy pickup – enabling the consumer to bypass notoriously long lines.
AI was certainly the big theme at last month’s Shoptalk conference. eBay CEO Devin Wenig declared in his keynote conversation that “If you don’t have an AI strategy, you’re going to die in the world that’s coming.” But much like omnichannel, the term AI can mean many things to many people. And it has quickly become a catch-all phrase for other technologies and approaches, causing confusion in the market. So what really is AI?
Breaking Down AI
AI is a science to develop a system, machine or software that can mimic human cognitive behavior, such as learning and problem solving. There are multiple components of technologies and processes that make up AI – data mining, machine learning and natural language processing to name a few. An application of AI in retail that most people are familiar with is Amazon’s Echo (Alexa) technology.
Machine learning, on the other hand, is a specific process within AI that refers to the science of self-learning algorithms. At its core, machine learning is about the use of statistics to solve problems using the data from the knowledge discovery process. It is also one of the most misused words when it comes to AI.
Eric Colson, chief algorithms officer at Stitch Fix agrees. Here’s what he had to say to Glossy’s Hilary Milnes at Shoptalk: “Most of the industry is rightfully confused about what is actually AI. Most of what we do is machine learning, but we do use AI, which comes into play on Pinterest. Our customers are looking at images on Pinterest and saying, “I can’t put this into words, but here’s what I like.” We take that and use AI to find similarities and visual aspects, and that results in better recommendations. That’s really just deep learning. We can also use AI to understand written reviews. One top got more reviews than the written works of Shakespeare, and it’s really hard to parse that stuff. With AI, we can understand who should receive that top in the future and who shouldn’t.”
Machine Learning in Retail Marketing
Robotic store associates and virtual baristas aside, there are very practical ways retailers can apply AI and, more specifically machine learning, to retail today. For instance, machine learning can help marketers understand the dynamics of each customer – not just based on previous purchase behavior, but by also learning how their tastes, moods and behaviors change and evolve over time.
By understanding the dynamics of each customer, retailers can more easily move shoppers from Point A to a more optimal Point B. If you have a subset of customers that always shop one category, such as ink and toner, and you would like to get them to also start purchasing their break room snacks from you, machine learning can inform you which marketing messages to deliver to that specific consumer – or group of consumers – over time to move them in that direction. Taking a machine learning approach to marketing is already helping early adopters maximize the frequency with which customers purchase, while not triggering the consumer fatigue created by rear-view mirror-oriented recommendations delivered by traditional predictive analytics and personalization solutions. This early success is quickly moving AI, and more specifically machine learning, into the mainstream discussion and related hype cycle.
Are Buzzwords a Bad Thing?
I’ve seen several Shoptalk attendees comment that just because something becomes a buzzword, it doesn’t mean it’s a bad thing or doesn’t work. The important thing with AI is understanding all its flavors and how each can help your retail business.
From optimizing internal business processes to understanding consumer sentiment and mapping customer data, AI can help retailers maximize the lifetime value of their customers across all touchpoints in the customer journey.
For more AI buzz at Shoptalk, we recommend checking out these posts: