All Customers Are Polygamists and How Direct Marketers Can Understand Them
Polygamy and consumers may not seem to meld in marketers' minds, but Peter Fader explains why they perhaps should. Polygamous consumers and the coin toss that is the wont of the desirable, inertial loyal customer are just a few highlights of a wide-ranging conversation with Fader, who co-directs a global marketing research initiative, the Wharton Interactive Media Initiative (WIMI), a "data-driven research effort on interactive and digital media."
Fader's also a direct marketing thought leader in his own right, which is highlighted in the Five-Minute Interview in the October issue of Target Marketing.
The following interview, framed by an explanation of WIMI, takes a look at the research done by Fader, a marketing professor at The Wharton School at the University of Pennsylvania. Fader also provides suggested reading related to this interview: "How Brands Grow: What Marketers Don't Know" by Byron Sharp and works by the late marketing researcher Andrew Ehrenberg, who espoused the polygamist consumer metaphor.
Target Marketing: You've seen the more specific questions I had for you; can you talk a bit about your research?
Peter Fader: … So I'd love to talk broadly about WIMI and then there's my own piece of it; my own research about customer retention and some of the paradoxes and so on that I've observed … But I think the WIMI piece is broader. I think it's potentially of greater interest, greater impact. So maybe we could start with that one. And, in doing so, I'd want to make sure that we bring in some of the other key members of the WIMI team, as well, to offer their perspectives. Because it's not just my show alone.
… I think you'll find, in talking to my co-director, Eric Bradlow and our research director, Elea Feit that they have their own unique perspectives that are aligned with mine, but still distinct. So that's one conversation. The other is, again, to dive into the research that I've been doing and some of the not only surprising findings, but some of the almost heretical findings about what we know about customer retention patterns over time and across people.
… I've had the luxury for 24 years on the Wharton faculty now of being able to choose what kinds of data sets I want to look at, what kinds of companies and industries I want to work with. … And so a couple of things have happened over the years. No. 1, I've observed that a lot of the basic patterns of customers doing things over time are remarkably similar, even when you look across seemingly unrelated domains. So a lot of the patterns that we'd see in a very mundane consumer packaged goods setting, you know, someone buying a kid's juice drink and deciding whether or not or when to buy it again and so on, would carry over surprisingly well to, let's say, more of a direct marketing, e-commerce kind of situation. It's just remarkable that when you strip away a lot of the nouns and adjectives that we use to describe our data sets and instead just say, 'You know what? We have an Excel spreadsheet and the rows are customers and the columns are some kind of activity over time. And we don't care if it's buying, we don't care if it's website browsing or contacts with customer service or whatever else. We just want to track customers over time, regardless of what domain it's from.' It's amazing how robust a lot of these patterns are.
So I spent a lot of time over the last 10 years, especially, trying to establish some of these similarities, some of these regularities and looking across different industries and saying, 'See? Here, it's happened again.' And occasionally saying, 'Hey, you know what? This one's a little different. This firm or industry's a little different. But at least we have some benchmarks to be able to be able to appreciate the existence, the nature and magnitude of some of those differences.
So, again, I've been lucky to have access to all of these data sets. But there's a lot of researchers who are doing really smart stuff. But they're in a vacuum and they can't reach out there and connect the dots with data sets as easily as I can or other folks at Wharton can. So we decided to set up this Wharton Interactive Media Initiative, less to focus on media or anything like that. We're not necessarily talking about the future of television, or anything like that. But more to serve as a matchmaker between companies that are sitting on tons of customer-level data and researchers, like ourselves, who thrive on it. … All we want to do is connect the dots and tell stories about customer dynamics and relate those stories back to the firms that are sitting on the data. … And firms come to us with interesting data sets and questions, we match them up with exactly the right researchers—whether they're from Wharton or anyplace around the world—and facilitate the whole research process, get the data in the hands of the right people, ensure that all the right legal and information technology safeguards are in place and make sure that the results of that research are communicated as clearly and broadly as possible. So that's what WIMI is, in a nutshell. …
TM: Your research seems to be concentrating on retention and loyalty marketing lately. Is this a business trend you're following that coincides with the downturned economy?
PF: It does coincide with it, but it really is a coincidence. It's not like I saw the economy turning and said, 'Oh, I've got to spend more attention looking at these issues. Companies need to tighten up their retention and loyalty practices.' It really is just kind of luck that the issues that I'm finding most interesting these days happen to be rising to the top of the agenda for a lot of firms: partially because of the economy, partially because of the greater availability of customer-level data, partially because of the different breed of employees who they're hiring to look at these kind of data; the different skill sets that they have. … But the patterns that I'm observing are, like I said earlier, are contradictory and often, like I said, heretical to a lot of the established wisdom. I'll give you a couple examples. …
And, first of all, let me preface it by saying, when we use the word, 'retention,' I'm very, very, very picky about the terminology. I'm referring specifically and exclusively to a contractual setting. One where a company knows when the customer goes away. You know when they don't renew their contract, when they drop their subscription, when they raise their hand and say, 'I'm out of here! I'm taking my business someplace else.' I don't think that the word retention applies to a transactional, a non-contractual firm. Let's say, like an Amazon. As much as I admire a firm like Amazon, they have no idea how many customers they have … So I'm talking specifically about situations where people are facing opportunities to renew their ongoing relationship. So I'm very specific about that word retention.
And so what companies have observed, and … this goes back for decades in direct marketing, is that retention rates tend to increase for a group of customers that stay with the firm for a longer period of time. So if we acquire a group of customers and we say, 'What percent of them stay with us from year one to year two? And what percent of them stay with us from year two to year three and three to four?' And so on. We tend to see those retention rates increasing as the cohort of customers ages. A pretty common pattern. There are rarely exceptions to it. But the issue is, 'What's that mean? What's that tell us about the customers?
And if you ask pretty much any company on the planet, they'd say, 'This means, that as the customers have a longer relationship with us, they like us better. They're feeling more loyal.'
Does that make sense?
TM: It does.
PF: But it's not true.
PF: And I've never seen evidence where it is true. And this is the heretical part of it. See, people draw the inference that, 'Hey, look. As this group of customers stays with us and their retention rate increases, that must mean that each one of those customers is becoming more loyal over time.
But that's not what's really going on. What's really going on is there's this shakeout process occurring that, when we acquire a bunch of customers, each one of them has a different propensity to stay with us. Some love us. Some don't. And what happens over time is we shake out the less loyal ones, leaving us with a different mix of customers, who tend to be the more loyal ones, right? It doesn't mean that the customers, that each individual, is changing over time. Just that the mix that we have is changing over time.
So the analogy that I like to use, and it's a surprisingly good one, is that each customer's given a coin and they flip a coin every time they have a chance to renew their contract. 'Heads, I stick around. Tails, I leave.' But everyone's carrying a different coin, some are more headsey, more inertial, more loyal; some are more tailsey, more flighty, more ephemeral. The coins don't chance over time. It's just that everyone flips their coin; the ones who are more likely to get tails leave and that leaves us with a more headsey inertial group.
So that story of how just the customer group shakes itself out over time turns out to be somewhat counterintuitive. It's easier to tell the story about how each customer becomes more loyal, instead of this shakeout process. But I've looked at the data from many different companies in many different ways and found that, pretty much, 100 times out of 100 it's the shakeout story that dominates the individuals becoming more loyal over time.
TM: What makes me laugh is it's the first time in a long time that I've heard 'inertia' used in a positive way.
PF: Yeah, and that's a great point. And we should be celebrating inertia. Most of what keeps our customers around isn't necessarily loyalty. It's not that they love us. It's just that they're too lazy to raise their hand and leave. And you think about how many people despise their cable company or their mobile phone provider. Yet even when contracts end … they stick around.
… Inertia is a wonderful thing and I believe, in fact, that it's a stronger force than true loyalty. Yet marketers don't like to acknowledge that. They like to think that they're doing things to make the customers loyal; to enhance the bonds and create more engagement and using all of these nonsense words, when, for the most part, it's just plain and simple inertia.
TM: Well, then how do you enforce inertia?
PF: You can't. And this is another heretical thing. And I'm overstating things a little bit to try to make a point—but not by much. So when we go out there and we go fishing for new customers, we throw our nets out there. We catch a variety of different fish. And, again, some are flighty, some are inertial. And the best we can do is hope that we get a pretty good mix of inertial ones. And, yeah, we could try to do things to try to make them more loyal. We could try to do a little bit of cross-selling, a little bit of upselling. We could try to get them to use our products a little bit more frequently; a lot of the standard kinds of customer-development tactics that companies like to use.
You've got to try those things. But the impact of those things is pretty marginal. And, in the end, what drives our success is just how good we were at throwing out those nets in the first place. The quality of the mix of fish that we caught in the first place. And, again, the quality is described as, 'How many inertial ones did we happen to catch?' And so we're better off thinking more about where to throw the nets next time we want to catch a bunch of new customers, as opposed to what we should do with them once we have them. So I believe that acquisition strategies will move the needle more than so-called retention strategies.
TM: Is this true, even though some marketers have been saying that actually, right now, acquisition is not the route to go?
PF: And I disagree with that. In other words, you're reading me exactly right. And, again, I just want to be clear. I'm saying things strongly to make a point. I'm not suggesting that retention marketing and customer development efforts shouldn't happen. It definitely should. You've got to be asking people, 'Do you want fries with that?' You've got to be recommending other products and services for them to try, in addition to the ones that they have. You've got to be doing all of those kinds of things. But the incremental impact that they're going to have is smaller than what a lot of companies think. And a lot of companies have been very disappointed by their cross-selling efforts. 'It's not working for me like it's supposed to.'
I believe that companies aren't spending enough on acquisition. I think that when they come up with their metrics of 'How much should we spend to acquire new customers?' CPA metrics, cost per acquisition, they're underspending. They're leaving money on the table.
TM: In "Customer-Base Valuation in a Contractual Setting: The Perils of Ignoring Heterogeneity," it sounds like you're warning marketers that they need to think of their customers as individuals instead of as an aggregated whole. What's the main change direct marketers need to make in order to implement this advice?
PF: The whole point of that paper is that if we don't do our customer lifetime value calculations the right way; in other words, we do them the naive way that too many firms and academics do them, our CLV estimates are going to be systematically too low; which means that we're going to be willing to spend too little money on new customer acquisition; which means, by definition, we're not going to focus on acquisition as much as we should. And so, again, I'm not saying that companies should be spending 10 times as much; it's not really that dramatic. But at the margin, when you have this incremental dollar or, let's say, $1 million to spend on top of our current resource allocation strategy, 'Where should we be spending most of it?' I would say acquisition. That there's a lot of low-hanging fruit that companies are unaware of and a lot of fine-tuning that can be done to make sure that we throw the nets in the more, let's call them, inertial directions. And, again, that runs quite contrary. But I believe that I have the statistical evidence on my side as opposed to the anecdotal wisdom.
TM: I never thought I'd hear 'acquisition' and 'inertia' in the same sentence.
PF: Yep. But I really do believe it. And, again, this isn't just based on some theoretical model, it's not based on some absurd economic rational principles. It's purely driven by the data. And it's driven by an agnostic look at data. Let's let the data, let's let the customers tell us what the patterns are, as opposed to imposing some kind of framework on them that forces it in a particular direction.
TM: How does this relate to your more recent research, such as "Customer Retention Dynamics in a Contractual Setting: The Paradox of Increasing Loyalty"?
PF: And now, let me add the newest piece to it. … But so, let me go back to my coin-flipping story. I'm basically saying every customer flips a coin, 'Heads, I stick around. Tails, I leave.' And my first-pass version of the story is to say that while the coins vary across customers, they don't vary over time. I'm flipping my coin time and time again until it comes up tails and I leave.
And a lot of people don't buy that story. They say, 'How can that be true? Of course your coin is changing over time.' So we built this new model to allow the coins to change over time. To allow for some dynamics at the individual level. And do you know what we find? The coin becomes more tailsey over time in almost every situation. In other words, as you stay around longer as a customer, your propensity to leave increases over time. And that's the most heretical thing I can say. Because there's not a person on the planet who would believe that.
TM: Except for, maybe, Christopher Ryan and Cacilda Jethá, who just released a book about how humans are not historically inclined to be monogamous.
PF: You know what? You're right. There's actually a lot of people on the planet who agree with me, but none of them are in marketing.
And I'll give you another example where the same patterns apply. It's not only in relationships among males and females, but it's also relationships between firms and employees. The very same patterns apply there; which is to say, every year the employee flips a coin: 'Heads, I stay on the job. Tails, I leave and get a new one.' And a lot of firms believe that, 'The longer they stay with us, the more loyal they become.' But there have been a number of heretical studies over the years, in the last 20, 30 years that have shown the opposite. 'What's up with this? That employees who stay with us longer become slightly more inclined to leave?' And I think that relationship between firm and employee and firm and customer is quite the same. And so it's another one of these examples about establishing similar patterns across domains and that basically, the exact same model that you could apply to customers, to employees and to relationships. I believe that those linkages, what it says about us as a species is much stronger than what marketers [believe], the Kool-Aid that they're drinking. So it's great that you're aware of that and, again, it makes it easier to establish some of these more general patterns. But trying to get marketers to believe it, it's an uphill climb.
… A lot of the models that I've developed, I rarely use the word 'polygamy' myself; although I have. But very seriously, a lot of the best principles for these kinds of customer patterns, stem from a gentleman in England, a guy by the name of Andrew Ehrenberg. And he [used] the P word all the time and [wasn't] shy about it. So I think it's actually a pretty well-established metaphor.
In fact, believe it or not, as you called, I was building an Ehrenbergian model on some data set for some confectionary manufacturer. And it's directly building on polygamist principles; which is to say that customers have relationships with lots of different providers at one time. They like some more than others. But, basically, when it's time to make your next purchase, again you're flipping a coin or you're rolling a die to say, 'Which one of my partners do I want to engage in this time?' Some sides of the die are a little bit heavier than others, so they tend to lean towards certain providers more than others.
But it's not like they're in a relationship and then they switch to another relationship. It's a bunch of different relationships going on at once. And I am no expert on how this translates to other forms of relationships. So I won't even claim that I've read the book that you mentioned or anything. But I am not the least bit surprised that people will discover these same kinds of patterns in not only very unrelated domains, but in ways that a lot of businesspeople would find downright offensive. But I believe it.