Customer Valuation: An Interview with Peter Fader, Professor of Marketing, the Wharton School of the University of Pennsylvania

Peter Fader is a leading authority on customer analytics and the author of the best-selling book “Customer Centricity”.

Just about every CMO will tell you their top priority is growing topline revenue. Where they might differ is how they go about achieving that growth.

There are two prevailing schools of thought.

The first is that growth comes primarily from attracting as many category buyers as possible, even if most of them are occasional users who buy infrequently. The opposing side argues that the cost of going after everyone in the market is a waste of resources: it makes far more sense to simply encourage existing customers to buy more, more often. Brand loyalty pays off in the long term, they argue, because it is much less costly to retain a customer than to acquire one.

This debate has been going on for years with all the shrillness of an ideological shouting match. On one side you have the Ehrenberg-Bass Institute for Marketing Science led by the iconoclast Byron Sharp whose immensely popular book “How Brands Grow” debunked a lot of taken-for-granted marketing principles. In a groundbreaking paper he wrote in 2002, he declared, “when brands grow, they can expect most of their sales revenue growth to come from having a larger customer base, rather than from an increased buying rate”. He based his conclusions on the NBD-Dirichlet mathematical model of brand choice developed in 1984 by his mentor Andrew Ehrenberg.

Sharp’s polarizing views certainly contradict the equally fervent beliefs of loyalty proponents who feel that marketers should apply disproportionate effort to increasing the value of current customers over their lifetime. Probably the best known advocate is Frederick Reichheld of Bain and Company who created the Net Promoter Score. In his classic book “The Loyalty Effect” published in 1996 he famously wrote that “improving the retention rate by five percentage points doubles the profit margin”. He goes on to conclude that according to Bain’s economic modelling, “Revenues and market share grow as the best customers are swept into the company’s business.” He doubles down on that business case in his latest book “Winning on Purpose” where he introduces the concept of “Earned Growth Rate” which refers to the revenue growth generated by “Brand Promoters” as a result of increased sales and referrals.

Like most abstract debates in marketing the truth lies somewhere in between. Companies certainly need to spend money acquiring new customers, although that becomes more expensive over time as the pool of potential first-time buyers contracts. But companies also need to invest in maximizing the value of current customers to drive profitable growth. In fact, customers should be thought of as assets whose value appreciates over time. The tricky part, of course, is to find the right balance between acquisition and retention spending.

That’s where Peter Fader comes into the picture. The Wharton School Marketing Professor believes passionately in a “barbell marketing strategy” which involves using acquisition dollars prudently to go after heavy category users while at the same time doing everything possible to please high value customers. The right balance is determined by doing a bottom-up study of behavioural patterns within the existing customer base. This analysis can pinpoint exactly how much untapped revenue potential there is amongst the high value customers who are the most likely candidates to expand their relationship with the brand. He calls this process a “customer-base audit” which he describes in detail in his latest book of the same name.

Stephen Shaw (SS): You were a math major at MIT and then went on to get a doctorate in marketing. Marketers usually have an aversion to numbers. What made you want to pursue marketing as a career?

Peter Fader(PF): I didn’t, is the answer. I was a straight math guy until this professor came up to me while I was in undergrad and said, “You ought to get a Ph.D. in marketing.” And I said to her, “You ought to get your head checked. I’m not going into marketing.” But she was very, very persistent and persuasive, and she just wore me down.

SS: What made her suggest marketing as a career choice?

PF: A couple of things. One is the times. This was the early 1980s, and we were just starting to develop the kind of customer tracking capabilities that we have today. Her pitch to me was, “We are building the electron microscope of the customer. We are going to have the capability to tag, track, predict in a way that we could never imagine.” She was one hundred percent right. At the time I wasn’t sure what I was going to do, but I was spending a lot of time thinking about being an actuary.

SS: The exact opposite of marketing.

PF: You’re right. But her point was we can use the same actuarial models. Instead of asking, “How long is it going to take until you die, it’s going be how long is it going to take until you buy?” This fairy godmother of mine, her name is Leigh McAlister. She’s now a Professor at the University of Texas. She just had incredible foresight about what marketing would become.

SS: It is remarkable because database marketing had just crawled out of the crib at that point.

PF: Absolutely. And that was a big part of it: she looked at some of the things that were going on, database marketing, direct marketing, late-night infomercials, and saying those kinds of practices should be more rule than exception. And today, so many companies are doing that kind of performance marketing without recognizing the debt they owe to those old-school direct marketers. They think they’ve invented something new.

SS: You’re quite right. My father, as an example, worked for 35 years at Reader’s Digest. Those were the days of mainframe computing and yet they were doing predictive modeling way back then.

PF: And in many ways better than what a lot of companies are doing today. Because back then, they were more scientific about it. They would wonder, “What do these numbers mean? What is our hypothesis? What kind of experiment can we run to test that hypothesis?”. It was just much more thoughtful. Today we talk about data science. There’s just not a lot of science in data science.

SS: Let’s talk about your book, “Customer Centricity,” which won you a lot of fame in 2011. What inspired you to write the book at the time? And what accounts for its success?

PF: The book was written out of frustration with companies that weren’t embracing a lot of these methodologies. Like I said, I’m a math guy. I build models. And these models work really well. Our ability to forecast how many customers you’re going to acquire, how long they’re going to stay, how often they’re going to buy from you, they work super well. Twenty years ago, I’d be just yelling at companies, “You’ve got to try this stuff! It works! Just give it a try! Here, I’ll give you videos, and R code, and technical notes, and spreadsheets. Just try it, will you?”. And most companies would ignore me. They’d say, “We’re busy. We’ve got a job to do”, or “You’re an academic. You’re not in the real world. What do you know?” or “It’s all very technical. There’s a lot of math there.” People would find every reason to either reject my stuff or to push it way down in the organization: “Sure, there’s someone who works for someone who works for me, and she’ll deal with that stuff, fine.” And so that’s a big reason why I wrote the first book, to say to the C-suite: “The world is changing, and you’re not changing with it”. There was some snark to the book – a willingness to call out specific companies and say, “Here’s why you’re not as good as everyone says you are.”

SS: Like Nordstrom.

PF: Nordstrom, Starbucks, Apple, Walmart.

SS: Your message 12 years ago was that companies needed to make a greater effort at understanding and serving their most valuable customers. Today the thinking seems to be that brands need to create community around their biggest fans.

PF: For the high-value customers, we must create community. But at the same time, some people think “Let’s just build a community and money will come raining down from the sky.” No. You want to make sure you’re measuring it. You want to make sure you’re bringing the right people together. You want to make sure that you’re checking how valuable they are and how much this community investment is increasing their value or helping us to acquire customers who are more valuable than the ones we’d acquire otherwise. So, for me, it’s always going to come back to customer lifetime value.

SS: We have made a lot of progress in terms of companies accepting the idea that they need to focus more on the customer. Yet marketers today continue to throw a lot of money at acquisition. What accounts for this mindset?

PF: Oh, easy, easy, easy. A couple of things. So, number one, we’re very sensitive to costs. Thanks to companies like Google, we know exactly how much it costs when someone clicks on that sponsored search ad. We know exactly what it costs as someone goes through the funnel. And so we’re just really painfully attuned to cost. I’m not saying ignore costs. But I want to give equal attention to future value. I want to know that the projected value of a customer should be the same as it’s costing us to acquire them. That if we can make value as visceral, tangible, measurable as cost, that’s going to change the calculus right there. If we can focus a little bit more on quality instead of quantity, it might shift the balance away from just acquisition to the care and feeding of customers after we acquire them.

SS: There’s good growth and bad growth for sure. But there seems to be a disconnect between brand marketing and performance marketing with the performance guys winning because they produce immediate sales.

PF: I have a love-hate relationship with both camps. The performance marketers are very short-term oriented. And the branding people are the antithesis of that: they want to avoid accountability and measurement at all costs: “It’s all about the brand”. And while there’s some truth to that, it doesn’t mean we can’t measure it. So, I love the fact that they have the long-term perspective, as opposed to the short-term performance marketers. I love the fact that the performance marketers have that quantifiable perspective as opposed to the brand people. Well, let’s just create the best of both worlds. And I think lifetime value is the way to unite them.

SS: Which I believe was the focus of an analytics company you started called Zodiac and then later sold to Nike. Tell me how that came about.

PF: What we saw is that we needed to scale the CLV models we were building from academic grade to full commercial scale, as well as add some other bells and whistles. We thought “Don’t trust companies to figure it out on their own: let’s do it for them”. And so that’s what we were doing at Zodiac. We created a means to get the word out in a way that I could never do just sitting in my academic chair. And we created lots of buzz in the industry, lots of testimonials. We had a line-up of companies saying, “Hey, work with us next.” And ultimately Nike bought the company in March of 2018 – what an incredible testimonial that was: A company that traditionally sold boxes of footwear to Walmart and Foot Locker, saying, “We want to have direct relationships. We want to know who’s buying what and what other things we can surround them with.” Even today, five years later, the fact that companies continue to ask about Zodiac, even though it’s long gone, shows what a great move that was.

SS: Certainly proof that Nike is maybe the most progressive marketer in the world.

PF: I give Nike credit for going that next step beyond being a client to grabbing the whole thing and embracing it, and not just the models but even the philosophical aspects of it. Again, very, very bold move on their part. And I don’t want to overstate cause and effect here, but you look at their performance over the years since they bought the company, every quarter beating investor expectations. Is it because of our thing? Nah. But it’s because of their own mindset, their own willingness to march to their own beat and picking up the skills that they needed along the way.

SS: Well, they’ve shifted away from retailers as a primary distribution channel and are pretty much committed now to going direct to consumer.

PF: Yeah, with their own shops as well. That’s right. And even when they do work with retailers, and they still do, the way they manage those relationships, the way they measure them, it’s just a very different way of operating than it had been. Unfortunately, it’s still more exception than rule, the way that they’re operating. Instead of every company saying, “We’ve got to do the Nike thing,” a lot of people say, “Well, that’s Nike, they’re different.”

SS: Five years ago, you started another company, Theta. That company, as I understand it, is designed to help companies come up with corporate valuations for M&A purposes, in part, using CLV, if I understand that correctly.

PF: So, I’ll tell you the back story. At Zodiac, most of the companies that we were working with wanted to use CLV to enhance and measure the effectiveness of their marketing tactics. But one of our clients was a private equity firm out of LA, and they didn’t care about any of the tactical stuff. They just wanted to know that when they’re thinking about buying that digitally-native women’s accessory company, what are they actually worth? And we can project how many customers they’re going to acquire, how long they’re going to stay, how often they’re going to buy, and how much they’re going to spend. Add that up, that’s the value of the company. My co-founder and co-author, Dan McCarthy, had worked at a couple of different hedge funds before coming back to Wharton to get his Ph.D. The guy’s super smart. It was Dan who was uniquely positioned to take all the goodness of the models, elevate them even higher, and figure out how to build the bridge to finance in a way that, not only would the models work, but that we could speak about it credibly, speaking their language and addressing their issues, their desires, their limitations, and have our models fit their needs. Dan has done that superbly well. And that just opened up all kinds of opportunities.

SS: Concepts like customer equity, brand equity, are usually buried under goodwill on the balance sheet. Is part of your mission to change how finance reports on CLV?

PF: The answer is yes. And it made sense until Dan said, “No, no, that’s just not going to happen. We should not be putting forward-looking projected numbers on accountable financial statements because they’re not accountable.” So, instead, let’s come up with accountable auditable measures that would be very tightly associated with lifetime value and customer equity, but things that we really could measure and report in a standardized way. What kinds of easily observable metrics could we have at our fingertips that would be a strong proxy for them? Let’s do so in a way that’s going to meet the conservative standards of accounting, but at the same time be strong indications of how much gas you have in the tank. And it’s been great. We’re actually getting public companies to start disclosing some of these customer metrics, which can be turned into forward value.

SS: On your web site you use the actual example of Warby Parker to suggest that they may be overvalued.

PF: Yes, that was at the time of their IPO. Fast forward a year and a half, and now they’re grossly undervalued. The whole point to that analysis is let’s not worry about stock prices. Let’s just worry about the unit economics, the value of customers. And if you think about it, the way that people buy glasses is pretty much the same today as it was a year and a half ago. The unit economics of a given customer, or the mix of their customers, is pretty much the same as it was back then. The value of the company hasn’t really changed very much. And so, these estimates that we come up with are not only diagnostic, and interesting, and ultimately accurate, but they also tend to be much more reflective of actual customer behavior, and often much more stable than the whims of Wall Street.

SS: The sorts of modeling you’re doing relies on fairly granular transactional analysis. What happens with companies that aren’t as data rich? Do you have workaround solutions for them?

PF: Absolutely. I am totally fine using different kinds of proxy measures to get going. NPS can be very useful in that regard.

SS: Should share of wallet be a beacon metric?

PF: Nah. I have nothing against that metric. The problem is, even at the individual level, it’s still lumping together a bunch of different behaviors. If your share of wallet as a whole is leveling off or decreasing, is it because customers aren’t staying as long, not buying as often, not spending as much when they do? I want metrics that help me single out and project one of those behaviors to get a more accurate, more diagnostic revenue projection.

SS: Is it your argument that CLV deserves as much attention as other KPIs like market share and share of wallet?

PF: Actually, no, no, no. So, CLV is the North Star that pulls everything together. Instead let’s report the things that will help us understand the components of CLV like customer retention and repeat buying. What percent of our customers who did something with us last period are still active? The other would be, among active customers, how many purchases on average did they make with us? How often are they buying? And how much are they spending? What’s the average revenue per user? As much as we talk about lifetime value, when we’re doing customer-based corporate valuation, we’re rarely doing that with lifetime value. We’re going to calculate lifetime value and we’ll report that to you – but the main thing that’s driving the valuation will be that next level down related to retention, repeat purchase, and spend.

SS: I want to ask you about Byron Sharp’s doctrine that market penetration trumps loyalty. He believes that market growth only comes by attracting as many buyers as possible, no matter whether they’re light or heavy category users. What’s your perspective?

PF: Well, let me first start by saying I am probably the strongest advocate for Byron Sharp and the Ehrenberg-Bass Institute down there at the University of South Australia. I’m probably the strongest advocate in the entire Western Hemisphere.
SS: You teach it, right?

PF: I teach it, hours and hours of it. I agree with focusing on penetration, focusing on getting the message out there broadly, focusing on a variety of different benefits instead of owning a niche part of the market. That is, I agree for 80% of the customers. Because they’re not going to stay that long, they’re not going to do much, and there’s not much we can do.

SS: They may not even be your customers.

PF: I love that point. And something that we emphasize a lot in the new book. For the so-so customers, which constitute most of your customer base, I’m 100% in agreement with Byron. But he’s wrong on the other 20%, which is to say he grossly underestimates the value of the high-value customers. And it turns out that the basic model that’s at the heart of what Byron does, which I believe in, too – the NBD-Dirichlet multinomial model – it’s a wonderful model, but it’s missing one component, which is basically how customers change over time. It’s a static model. And when we bring in what we call nonstationarity, allowing customers to evolve over time – something which, by the way, Byron’s mentor, Andrew Ehrenberg, knew full well – that when we bring that one piece into the model, and it’s important, because then those high-value customers actually become even more valuable and more important. And that’s where all this focus on the right customers for strategic advantage comes in, that we make sure we’re doing the Blue-Ribbon Club for those high-value customers, but recognizing there’s very few of them. Incidentally, Byron and I have just a wonderful relationship. Almost all of our exchanges are very, very positive, even if we disagree. But the high-value customers are different and we need to do different things with and for them.

SS: A CMO typically gets 10-12% of the company’s operational budget to work with and then invariably spends 80% of it on acquisition and 20% on customer retention. Based on your thinking, shouldn’t it be the other way around?

PF: Well, yes and no. It all depends on how that acquisition budget is being spent. If it’s being spent, as it is by most companies, on purely performance marketing – let’s go after as many customers as we can, as cheaply as possible – that’s a problem because we’re going to acquire a whole bunch of customers hoping and praying that they become good. They probably won’t. If instead, we’re spending that budget a little bit more on quality instead of quantity, then it might not. Now, if it’s 80% on acquisition, 20% on retention, that is a little too imbalanced, I’ll agree. But it’s not so much the overall quantity of dollars that companies are spending on acquisition, it’s how they’re spending it. And that tends to be what’s more troublesome, this hunt for low-cost volume as opposed to high-cost quality.

SS: The big challenge for marketers is showing the correlation of that spending to the company’s bottom line. It seems that’s where you fit in.

PF: Exactly. And, for me, it’s going to be these same basic building blocks I keep talking about: acquisition, retention, repeat purchase, spend. Now, it’s not enough just to have metrics around them. We need to understand the reasons why. And that requires more qualitative measures. The problem with many companies is they’ll do the qualitative stuff across the whole customer base. I say let’s do it separately by, say, lifetime value tiers. Let’s look at our top 10% of customers and see what they’re saying, what they think, how they feel. What are their needs, desires, frustrations? And how is it different from tier number two, or three, or four, or five? So, let’s figure out the emotional triggers for each type of customer, each value tier of customers, rather than trying to do it on an overall basis.

SS: Should customer accounting be an independent department, operating as a bridge between finance and marketing?

PF: Bless your heart. Such an important question. Too often we either neglect or greatly downplay the role of costs in the equation. Let’s make sure that we have all of the costs accounted for. So, yes. It’s really, really important to do it right, to do it conservatively, to do it in an auditable manner, and to weave it in with all of the revenue and value metrics that we’re coming up with. We’re getting really good on the revenue and value side, but it’s still pretty messy on the cost side.

SS: I’m glad to hear that you want to crack the code on that somehow.

PF: Yes. There should be big, obvious, transparent, agreed-on standards. We need to have conversations with some of our accounting colleagues. I think there’s a lot of work that needs to be done. We don’t want the people in accounting or finance to be looking at us as just a bunch of lightweight marketers. We want them to say, “You know what? You really do have something to contribute that’s of value to me.” And I think we’ve been making a lot of good progress in that regard.

Stephen Shaw is the Chief Strategy Officer of Kenna, a marketing solutions provider specializing in delivering a more unified customer experience. Stephen can be reached via e-mail at sshaw@kenna.