Analyzing customer data to predict future purchases can be hard. And it doesn’t help that most companies focus far too often on preference data to anticipate what customers will want next. Companies used to struggle to find the tools and techniques to engage customers and create personalized buying moments, but today most retailers can deploy technologies that track purchases, understand patterns and learn customer preferences. In fact, analyzing preferences has become something of a point of pride for many retailers.
They talk about hyperpersonalization. They analyze their custom-made dashboards, seeking insights that will drive purchase behavior. Retailers want to know, among other things, why people abandon shopping carts and why people choose only one item when additional items are suggested.
The how matters. And I’m not especially impressed by what most companies call preference data. In fact, I think the last 10 years have shown that it’s not preferences that drives purchase behavior, it’s situation.
The traditional argument for customizing a purchase experience goes like this. Customer loyalty improves when retailers use customer preference data to customize a buying experience. For example, if someone buys a blue, cabled sweater, then they become a blue, cabled sweater type. And by looking at what other BCS (blue, cabled sweater) people types like, the company can recommend other cabled sweaters that the customer will like. That’s preferences analytics.
Advertisement
But what the last 10 years — and especially the last five years — have taught us is that the primary reason people buy is because of a change in their situation. Certainly COVID-19 and the shutdown drove purchase behavior. It’s the change in situation that creates the market for the purchase. The most obvious example of this phenomenon is seasonal purchases. Right?! When the weather changes, people’s needs change.
So what actually drove the purchase of the blue, cabled sweater? Was it preference or situation? Did the weather change? Did a fashion trend shift opinions about 1980s garb? Did the purchaser buy the sweater as a gift for someone else? If any of these scenarios are the real reason for the purchase, then all bets are off on the ‘preference analytics’ retailers use to anticipate what the customer will want next.
If retailers want to get the next purchase right most of the time, they need a new model of what drives transactions. Here’s what that model looks like: When a new situation arises people have new needs. The best way to anticipate a customer need is to comb purchase data for insights into new situations that are arising.
Back to blue, cabled sweaters — which I think I owned in the 80s. The question that the analyst should be asking of the data is: what are the situations where BCS’s are the perfect solution?
People Get Into Modes
Let’s take this insight — that situation drives the purchase further. Consumers recognize that they respond differently to different situations. Over the last 10 years, we’ve studied how people respond to situations and new technology. Their overwhelming behavior: they get into modes.
A mode is a mindset and set of behaviors that people get into temporarily. People enter different modes because these modes (e.g., learning mode, exercise mode, etc.) help them maximize their impact (e.g., effectiveness, productivity, etc.) in various situations they encounter.
Sometimes, getting into modes is a passive response to situations, such as procrastination mode, but frequently it is a proactive activity, such as ‘get-it-done mode.’ People we’ve studied take time to think about and plan their modes. They talk openly, often on social media, about the modes they are in and why.
To understand a customer’s mode is to be able to understand much about their situational needs. If someone is in planning mode, you can recommend products that will support that situation. More often than not, the millions of people who get into planning mode will have similar needs.
Because common modes (like ‘parent mode,’ ‘work mode,’ ‘vacation mode’ and ‘beast mode’) are, well, common, retailers can prepare for them. They can study modes. They can use their data to identify potential opportunities to support their customers. Some modes are so obvious that most retailers already have a plan. Think of ‘back-to-school mode,’ ‘weekend shopping mode,’ and ‘new home setup mode.’
Blue-cabled-sweater mode might be exactly what the customer was thinking when he or she bought the sweater to begin with. Or, put another way: the sweater was purchased for a particular type of situation — not because the customer plans to redo his or her wardrobe with sweaters.
Wouldn’t it make more sense for data analysts to focus their algorithms on modes and situations rather than assuming that every purchase indicates a preference?
Dave Norton, Ph.D. is the Founder and Principal of Stone Mantel, a research-led consultancy at the forefront of customer and employee experience strategy. For the last 20 years, he has helped companies like Royal Caribbean, Marriott, US Bank, Best Buy and Clayton Homes set new standards for customer, service and employee experience strategy. Alongside Aransas Savas, he hosts the Experience Strategy Podcast; a show that explores how organizations create deep and meaningful relationships with their customers in a way that delivers on Time Well Spent, Time Well Saved or Time Well Invested. He received his Ph.D. in Rhetoric from the University of Minnesota and currently lives in Salt Lake City.
