At Jornaya’s Journey 2016 client summit, taking place October 25-26, attendees will have the privilege of hearing the keynote from Peter Fader, Professor of Marketing at the Wharton School of University of Pennsylvania. Professor Fader was kind enough to chat with me recently to answer a few questions to provide a preview of his keynote.

Q1. How can marketers use behavioral data to understand the shopping activities of their customers?

First of all, it’s not just what you do with behavioral data—but it’s the fact that you are using behavioral data, as opposed to other, much weaker sources of data. So, many marketers who have very rich behavioral data at their fingertips don’t start with it or don’t use it at all. They are still so obsessed with things like demographics and other measures that are so easy to observe but aren’t necessarily indicative of what consumers are really going to do. While these other data sources do have some value, behavioral data is at the top of the list.

It may sound corny, but it’s true: Nothing predicts future behavior better than past behavior. And that’s what we want to know—who is likely to do what and for how long and for how much money? And, what can we as a brand do to get more? What you really want to do is start with the behavioral data and squeeze as much value as you can out of it before you start layering on the demographics and the social and the biometrics, etc.

It’s more than just taking that behavioral data and just simply summarizing it. These days so many people/companies are really into data visualization and data science, which is great stuff. But, very often data science/visualization ends up being just pretty ways of looking at the behavioral data. What you really want to do is to recognize that what we observe people doing is a decent but imperfect indicator of the true underlying propensity to do things.

For instance, the fact that you bought a widget three times this year doesn’t necessarily mean you are going to buy that widget three times again next year. But, it does tell me something about whether you are a heavy buyer or a light buyer. And, if I have enough of a history (and, perhaps other relevant data inputs) I can start to get a picture of you that’s much more forward-looking (and actionable) than mere summaries of the past data. That’s what you need to care about. That’s what you want to segment your consumers on. It’s not what you have done, but what you are likely to do next.

Get beyond obsessing over the data, per se, and use the data as fuel and ingredients to tell stories about people’s propensities.

Q2. How crucial do you think it is for marketers to have a good understanding of their customers’ shopping behaviors, and how can they leverage that knowledge in their marketing efforts?

Of course I’m going to say it’s crucial and there’s no doubt that it is. But, how you gain that understanding and what you do with it depends on what your objectives are. For instance, you might want to know which of your existing customers are most likely to churn, or who among a set of your prospects are most likely to convert into customers in the next month. If you want to make these kinds of short-term classification statements, than traditional machine-learning methods are great. But, if you want to make longer-term assessments, like customer lifetime value (CLV), where you’re not so much worried about tomorrow, but you’re worried about long run overall propensities, it’s a very different set of analysis and questions that you’re going to ask.

The good news is that the behavioral data you can collect can lends itself to many different action-oriented applications, like those just mentioned, among many others. But, it’s not just a matter of saying well, let’s run “the model” (whatever it might be) and smart things happen. You need to be very thoughtful about what different methods and different uses you’re going to get out of the data for the question you are asking at this moment.

You really need to have a broad variety of methods in your toolkit to address the different kinds of questions. Unfortunately, a lot of people begin and end with machine learning. There’s no doubt that machine learning is definitely one of the major components of our tool kit. But, a lot of people stop there and think that if they can master machine learning that they are all set. But, while it’s undoubtedly powerful tool, it’s only one of many skills that a data-driven company should master.

Q3. How can a consumer’s behavioral data inform a marketer about the level of buying intent of that consumer?

It’s surprisingly simple. This exact question has been asked for many years. Companies have been asking “What is it about someone’s past that is most indicative of their future value?” It is a fundamental question, and therefore it kind of makes me sad people don’t have the answer at their fingertips, because it’s easy.

The answer is RFM, which stands for recency, frequency, and monetary value. Before diving into what that means, let’s back up and let me talk about a history lesson because our forefathers in direct marketing were asking this same question 50 years ago. So, they decided to look at which bits of data from the past are most correlated with—and therefore presumably predictive of—the future. And, this is how they found RFM. It wasn’t through any fancy statistical model or through any experiment. It was just seeing which bits of data they saw correlated over periods of time. It’s a simple way to discover it, which is why it makes me sad because it’s so true in so many domains.

I’m just amazed that if you are looking at so many different business settings—from retail to mobile gaming to travel services to prescription refills—you just give me RFM, tell me the last time someone bought it from me or how many (recency), and tell me how many times they bought it over the past two years (frequency), and what’s the average size of the purchases (the monetary value), and I can give you a really accurate statement about what any customer with that customer’s value will be over the next three (or more) years.

And, it doesn’t stop there. RFM doesn’t have to be built on transactions, per se. Recency might reflect the last time was that someone visited your website. Or referred someone to your business or tried a sample. It’s the same with monetary value. We usually think about it as dollars spent, but it could also be some degree of engagement—”How long did they spend on our website?” “How much have they used our content?”

RFM is a pretty broad framework, it’s so powerful, it works so well, and it’s so easy. Yet, I’m going to guess that most of the folks who are reading this blog will not be familiar with RFM.

As a professor, it’s my job to find best practices from the past and shine a light on them and say “you don’t need to reinvent the wheel”, and at the same time pushing the frontier out and saying that, assuming you’ve got this message and you’ve mastered the RFM, and you’ve squeezed all the value out of it, now what? Then you need to take the other data sources and metrics to go even further.

Q4. What kinds of variations and similarities do you see in shopping behavior across different industries/segments?

Indeed, I spend a lot of time thinking about this question. The usual ways we divide up different industries—B2B vs. B2C, product vs. service, international vs domestic— a lot of those distinctions are not really that big or that interesting when you get down to the level of the data.

People are surprised by that; they can’t believe B2B and B2C would have buying patterns that look the same, but they do. But, if you put your stereotypes aside and look at the patterns, look in an excel spreadsheet and look at a bunch of customers in the rows and here’s their activity over time, just look at the data.

To me, the biggest distinction is whether it’s a contractual or a non-contractual relationship. When talking about RFM before, I was referring to a non-contractual relationship, which characterizes most businesses. Most times we buy things now and again, come back for a while, and at some point no longer need or want the product or don’t want to work with this vendor. Or, we move or die or some other reason why we drop out. Whatever the story, it can be summed up that we buy occasionally for a period of time, then we stop using it.

So, most of life is characterized by this idea of “latent attrition”. This is the idea that you don’t tell your grocery store manager, “Okay, I’m not going to come back here anymore.” You just stop shopping there.

So, a really big challenge for marketers—and few of them verbalize or recognize it—is both anticipating and preempting the idea of latent attrition. It’s not a matter of waiting until the customer has been gone for a long time and then have to figure out how to win them back. It’s the idea of how you can look at their patterns and say, “You used to buy from me, but it’s been awhile since you’ve around, and that’s telling me something”.

The non-contractual setting, which is very common, is also very challenging because latent attrition is a big part of it. Meanwhile, on the contractual side you have situations where customers self-identify that they are not renewing your contract or subscription. More and more companies have moved in that direction because they want their customers to be engaged on an ongoing basis and they want to have those insights into when their customers are possibly leaving.

So, that’s a really big distinction right there, where we will build separate models in each case.

Q5. Your bio states that you believe that marketing should not be viewed as a “soft” discipline. Can you expand on that?

Yes, it breaks my heart when I walk into an ad agency and they kind of size me up and they ask, “so, are you a creative guy or an analytical guy?” And I’m asking, “Why can’t I be both?”

So much of marketing for so much of its history has been about just what color should the packaging be or which celebrity should we get for our Super Bowl ad? And for a long time it was all about that stuff and that stuff alone. The ability to package your product the right way and have the right message for it and have the right spokesman for it is all important stuff and there’s real skill required there, don’t get me wrong. But, marketing doesn’t begin and end with it.

Being able to look at the behavior and predict who is likely to do what and then bringing in the creative part to get them to stay longer and buy more is a beautiful interplay. But, as a field we’ve been so imbalanced towards one side or the other, so there’s actually now a lot of suspicion where the creative types will look at the the analytic types and say, “oh, they don’t know marketing – they’re just egghead number-crunchers.” And, both sides are guilty of this. We need more mutual understanding and respect and to raise the visibility and respect for marketing across the organization, we need to sync up between the soft and hard activities we perform.

Q6. You are going to serve as keynote speaker at Jornaya’s upcoming Journey 2016 summit on October 26; what can we expect to hear about?

For the most part I’m a number cruncher, but a creative number cruncher. But, that’s not enough. In recent years, I’ve put a whole strategic framework on top of CLV—that’s theidea ofcustomer centricity. Basically, if we can use data to find out who our best customers are, and if we can find ways to enhance their value and find more customers like them, we can make more money than if we just focus on version 2.0 of the product.

It’s all about finding and creating value out of our customers. And, what I’m going to talk about at Journey 2016 isn’t just that philosophical idea, but, I want to get specific about how to make the money off of this philosophy. I will talk about the specific ways that this customer-centric perspective can lead to increased profits. I’ll use real-world examples and talk about perspectives and methods that a lot of product-centric marketers would probably disagree with. I will take a different strategic view and talk about how customer centricity can be profitable. I am really looking forward to not only putting these ideas out there, but also to hearing the feedback from the attendees.