Too often, companies aren’t getting the most value out of their customer behavior analytics. With the advent of big data, companies have more data than ever about them — and less actionable insight. In today’s episode of “Datameer Whiteboards,” I’ll walk you through how customer behavior analytics can help you and what it looks like today.
What if you could understand what makes your customers tick?
That’s what customer analytics is all about. It helps you understand what motivates customers and what drives them to take the next action in terms of how they interact with you. You can use this information to maximize the relationship you have with those customers. The digital age has dramatically changed the way customers behave.
Did you know that 56 percent of all customer journeys involve multiple channels and involve multiple events?
Big data analytics can help you understand this customer journey. You can understand the behavior behind that journey to help build a better relationship with your customers. By understanding behavior, you can also identify the next-best set of actions to take with a customer and therefore, you can help drive far more valuable outcomes. It’s about talking to the right audience at the right time with the right message.
When you’re creating customer behavior analytics, you usually do it in four steps. The first thing you do is you go and gather all the data you have about customers, whether it’s how they’re interacting with you on a digital front, what transactions they may be making in physical stores, or ways to know better information about the customers themselves such as demographic information. You gather all this data together so that you can use it in your analysis.
The second step is to try and identify behavior patterns within that mass of data, whether it’s coming from digital channels or non-digital channels. You want to understand what journeys they’re taking, what behavior patterns they’re taking. Once you’ve done that, then you want to take a look at groups of customers that have common characteristics that follow these various behavior patterns so that you can identify these customers and be able to understand how to target them. The last thing you do is you group, or classify, all of these customers into these segments so you know how to treat them at various points in time in the customer lifecycle.
Speaking of the customer lifecycle, well, customer behavior analytics help you in all four phases of it.
Here’s how customer behavior analytics can help you at each of these phases. First, in the acquisition phase, they can turn around and tell you what best offers does it take to get a customer. They can also tell you what channels do I need to reach out to to find the most profitable and best, and most loyal customers. They can also help me to understand how to lower my cost of acquiring customers.
When I go to the next phase of engagement, they help me understand what cross-sell or up-sell offers I need to make to customers. Therefore, help to try to drive more activity with those customers. They help me also understand how I can increase overall purchase value, or increase the rate at which customers purchase from me.
When I reach the retention phase, they help me identify patterns that might lead to churn, so that I can identify when these patterns are occurring and help reach out to customers to give them next-best offers to keep them in the family. Lastly, in the loyalty phase, what I’m trying to do is trying to better understand how I can motivate customers to stay more loyal. What programs I can offer to them that help them purchase more from me and maintain a very long lifetime value with that customer?
The old-school way of doing customer behavior modeling followed a process called recency, frequency and monetary value. What that meant was customers who purchased most recently from me were most likely to purchase again. Customers who purchased the most from me, or the most often from me, were more likely to purchase again. Customers that spent the most monetary value with me were also likely to come back and purchase again. That was the typical behavior model, but then you could only really understand transaction data. That’s the way this model was created. That’s an old school model that doesn’t really apply in the digital age.
In the digital age, a new model has appeared. It’s multi-channel because you have many different ways in which you’re going to transact and interact with a customer, whether that’s online, or mobile, or physically in stores or branches, in a number of different places.
It is detailed which means you want to look at a very granular level in order to understand how I can personalize the marketing to that particular customer. It is time series, meaning I need to look at how the customer’s behaving with me over the course of time so I can get a better gauge of what next action they may take. It is also about the lifetime value, or maximizing the lifetime value, of that customer. Rather than looking at recent purchase history, I want to better understand and potentially predict what type of lifetime value I might get out of that customer based on the next-best action I may want to take. I’ve used this word next-best action a couple times already. What that refers is I’m using all this information to derive this next best action in order to best figure out how to market and motivate this particular customer.