Predicting Customers and Transforming Markets: A Conversation with Professor Peter Fader

00:07 Dr. Fader

For me, a lot of my career here at Penn is paying it forward.

And that's why I just love being involved with CURF and related organizations because there's a lot of smart kids here. Some of whom want to do research, and that's great. But there's many who, for whatever reason, like me, don't think it's for them. You know, a variety of different reasons, social pressures and so on. And so for me to show them the fun, to show them the opportunities, to give them the variety of careers above and beyond going to Wall Street, I view that as an obligation, as a pleasure. And I'm just so thankful for all the undergraduates that I've worked with and continue to do so.

00:59 Nora

You're listening to Ever Thought About...?, created by undergraduates at the University of Pennsylvania. We hope to bring you exciting episodes about the diverse research undertaken around campus. Sit down with us as we chat with Penn professors about the work they've dedicated their lives to.

In this episode, we explore research on marketing with Professor Pete Fader. He is the Francis and Pei-Yuan Chia Professor of Marketing at the Wharton School here at Penn. His research focuses on analyzing behavioral data to understand and forecast customer activities. He works with firms across many different industries and has even co-founded several companies to apply his research. One of which, Zodiac, was acquired by Nike in 2018.

Professor Fader has dozens of publications in top journals and has won numerous teaching awards, including the Lindback, the most prestigious teaching prize awarded at Penn. He's also the author of three books: The Customer-Based Audit, Customer Centricity, and The Customer Centricity Playbook. He has been featured in The New York Times, The Wall Street Journal, The Economist, The Washington Post, and on NPR, among other media.

In 2017, Professor Fader was named by Advertising Age as one of its inaugural 25 Marketing Technology Trailblazers and was the only academic on the list. He has taught many innovative courses, but today he focuses primarily on applied probability models and marketing, a very popular and challenging quantitative elective.

Welcome, Dr. Fader.

02:19 Dr. Fader

It's great to be talking to you, Nora.

02:22 Nora

Awesome. Thank you for being here. So that was a really impressive and long introduction of you that I just read. But I kind of wanted to go back all the way to the beginning, or maybe even before the beginning of your career, just to start us off. So I see that you graduated with a degree in math from MIT. How did you end up venturing into marketing research?

02:42 Dr. Fader

It's a crazy story. Yeah, I was just a straight-up math major at MIT. I liked crunching numbers, playing around with data, and forecasting things. I figured I'd end up on Wall Street. I was even talking to the NSA about cryptography, all sorts of crazy things that a mathy person would do.

Then this one professor at MIT, she was at the Sloan School, a marketing professor. Her name is Lee McAllister, she's at the University of Texas now. She came up to me during my senior year. I'd just taken a course with her, and she said, "You ought to get a PhD in marketing." And I said to her, "You ought to get your head checked. I'm a math guy. I'm not going into marketing."

But she was very persistent, and very persuasive, and very compelling. And she painted this picture—keep in mind this is 1983, before Mark Zuckerberg was born, probably—and she said, "We're building the electron microscope of the customer. Pretty soon we're going to have the ability to tag and track, and all that mathy stuff you like to do, we're going to have so much cool data for you. You’ll have lots of opportunities to do it. No academic is doing this kind of thing. No company is doing it. You can set the path."

It was a little over the top, but not too far from it. And I thought, you know what, why not? I like crunching numbers. If half of what you're saying is right, then maybe this is an area to go into. I have to admit, through the PhD program at MIT and even my first few years on the faculty here, I was pretty skeptical. I was like, “how long am I going to do this marketing thing before I get on with my life and actually use my skills?”

But then as the data started to flow, I realized—you know what? She is right. You know what? This is fun! And you know what? I really can make a difference. So I've been here 37 years now on the faculty, and every day just gets funner and funner all the time.

4:47 Nora

Wow. So she had that vision of how marketing was going to be shaped into.

04:52 Dr. Fader

She had a vision for the field. She had a vision for me. I still refer to her as my fairy godmother and continue to make pilgrimages to Austin every year just to thank her and seek her advice on the things I'm thinking about and working on. She was right all those years ago, so I might as well keep tapping into that great source of knowledge and advice. She's never steered me the wrong way.

05:18 Nora

Wow, that's really cool to have that kind of mentorship and research.

05:22 Dr. Fader

For me, it's been very much about gratitude for all the things she has done and continues to do. A lot of my career here at Penn is about paying it forward. That's why I love being involved with CURF and related organizations—there are a lot of smart kids here. Some want to do research, and that's great, but many, like me, don't think it's for them for a variety of reasons: social pressures, uncertainty, and so on. So for me to show them the fun, the opportunities, to give them a variety of career above and beyond going to Wall Street, I view that as both an obligation and a pleasure. I'm just so thankful for all the undergraduates I've worked with and continue to do so.

06:08 Nora

Yeah, that's awesome. Could you actually talk about some of the projects that you've done with undergraduates?

06:13 Dr. Fader

So I work on one hand on a wide variety. On the other hand, the stuff that I do, big picture, is kind of narrow. I like building models, looking at data to predict all sorts of things about customers: how many will we acquire, how long will they stay with us, how often will they buy, how much will they spend, how will that change under different kinds of conditions?

It's a surprisingly rich area of research. A lot of people would hear these kinds of things and just think it's sort of crass, mercenary work helping companies squeeze more money out of unsuspecting individuals. And I'm well aware—painfully aware—of some of the bad things that companies can do armed with this kind of data and knowledge.

But from a pure intellectual standpoint, these are very very interesting forecasting problems. It's very very interesting to combine a lot of the statistical methods that myself and a lot of people in marketing use with psychology. Let's take interesting psychological theories and figure out how to bring them to life statistically on real-world messy data, not some kind of nice lab experiment.

Can we actually find some of those effects? Can we find the limiting conditions? Can we understand what the implications are for managers?

That's the kind of work that I do. And with undergraduates, it can be everything from doing the math—solving horrible, nasty integrals to come up with a new model—to figuring out computational shortcuts so we can bring these models to life at full scale with millions of different customers. It also includes thinking through limiting conditions, running lots and lots of simulations to understand when these things work and when they don't, and so on, as well as building bridges to other fields—trying to take these models and bring them to seemingly unrelated domains. And finding methodologies from other fields and see if they can actually relate to customers as well as they do to, you know, planets moving or chemicals in a flask. I'm always open to new methodologies and new applications.

08:27 Nora

Wow, that's really cool. Could you talk a little bit more about the bridges to these other fields that you just mentioned? Like maybe not just business, but also law, medicine—and then you mentioned something about planets?

08:39 Dr. Fader

Well, it turns out that the work I do, building probability models, really goes back to some of the work I was doing as an undergraduate. Again, I was considering a lot of different career paths, and the main one I was thinking about was becoming an actuary—the people who work for insurance companies and basically set the rates.

It's very important for them to know how many people are going to live to a certain age and how that will vary based on whether they smoke, or jump out of airplanes, or ride motorcycles. What's interesting about actuaries is that they kind of know their limitations. It's really hard for them to say, “how old is Nora is going to be when she passes away?” But it's really easy for them to look at a group of customers or policyholders who share similar characteristics and say, “what percentage of them will live to be 85 years old?”

So basically, I'm taking actuarial models, viewing the world probabilistically, and making statements about how long it's going to take customers to buy instead of how long it's going to take them to die.

And it turns out that the same probabilistic methods both apply to—and were refined and developed in—lots and lots of other fields, including physics, including manufacturing reliability, and epidemiology, and so it's important to keep up on those literatures. It's important to read papers on wildly unrelated topics but with your eyes open to saying, "I'm going to find the connection here. I'm going to find how this relates to some aspect of customer behavior."

And again, turning around and saying, "Some of the methodologies that will develop can be surprisingly useful in other domains as well."

There is nothing more gratifying than when I hear from a researcher in some field that I've never even heard of saying that they've stumbled onto one of my models and applied it. Instead of customers, they might be applying it to, for instance, some animal tracking.

Or we're sitting right here in Van Pelt Library. My work has been applied to—I don't even know what we'd call the science—bibliometrics, to try to solve a very practical problem. You know, as libraries are getting squeezed for space, they have to make tough decisions about which books to keep in the library and which ones to put in remote storage. So they need to predict how often a book is going to be borrowed over the next year or two. And it turns out that the very same models we'll use to predict how often a customer will buy from us over the next year or two can be applied directly, with no changes, to these other kinds of domains. And again, it's great fun to see that. It's nice to know that the research has an impact above and beyond marketing.

And again, turning around and asking, “well, what can I learn from you?” Then coming up with models and practices that can just be the best of all worlds.

11:29 Nora

Wow. So it sounds like you mentioned that the research that you do in and of itself is sort of like a narrow field, but it sounds like it has so many applications outside of it as well.

11:38 Dr. Fader

And I didn't believe that myself. Like so many researchers, you tend to just look at what you do and think we're building methodologies just for our little tiny field. But part of it was the fact that I'm coming in as an outsider. So I've always kind of wondered what we could borrow from other fields like actuary science, and what we could apply to other fields. Like, I love sports statistics and things like that. So can we, you know, have anything to say about what's going to happen in Sunday's Eagles game, for instance? I've kind of had that curiosity to go beyond the usual barriers that define—and sometimes limit—a field.

And I have to admit, for the early part of my career, I was pretty narrowly focused, trying to get tenure as a professor in marketing. So you don't want to go too far afield. But these days, as an old full professor, it's really great to see those broader applications.

And I probably enjoy them more than I enjoy writing just another paper about, "Here's another thing about what customers do." But still, I do a lot of that stuff. That's the day job. So it's important not to give up on it.

12:50 Nora

Got you. So you talked about how in the beginning you had this sort of narrow focus on that path to try to get tenure and then you started sort of expanding outside of that. Could you talk a little bit more about some other ways that your research has evolved over time?

13:03 Dr. Fader

Sure. For a variety of reasons, most of the work I did as a junior faculty member—up until tenure—I was very lucky to get data from a company that collects data from supermarkets. So all the old papers I wrote were about orange juice, salad dressings, kids' juice drinks, and stuff like that. I had lots of data, and, you know, it worked. I got tenure. Yay!

And I was perfectly happy just kind of wallowing around in my consumer packaged goods sandbox. But then a couple of things happened in the 1990s. First, one of my students came up to me and said, "Why are you working on all these boring topics? Why don't you work on something interesting, like the music industry?" And we all love the music industry, well we all love music one way or another. So we set out to try to bring some of this data stuff to the music industry—going to record labels and saying, "We can help you predict the sales of that next album, and to sequence what the songs should be, and stuff like that."

This was the early 1990s, and the music industry wanted no part of it. They were just not interested in this kind of geeky academic stuff—until the late '90s came around, when all the digital downloads started, and they didn't like any of that. That became really interesting because it was another data source, to not only know who was buying which album and when, but also which songs they were downloading and all that sort of thing.

So I got really deeply involved with the music industry. In fact, you probably don't remember the original Napster—the original file-sharing platform that changed everything back in 2000. I was an expert witness for Napster in this big epic case when they actually got shut down. But it really opened my eyes to a broader set of applications. I realized, you know what, even though that was kind of an unsuccessful one, this stuff can really matter, and it doesn't have to be limited to kids' juice drinks and so on.

It was right around that time when the whole dot-com thing was starting. So we could start collecting data on, well, pretty much everything. We'd get data from all these ridiculous companies, things like pets.com. I remember there was this company back in the old days that used to sell books online. It was named after some river. What was that one? Amazon or something like that. 

15:30 Nora

Laughs

15:31 Dr. Fader

All these companies are long gone, but it was really, really interesting to get in there early, look at that really rich data, and to be able to see who was doing what, and come up with forecasts. It was right around that time that those kinds of companies were much more able—and therefore interested—in coming up with forecasts and building their businesses around them. So, the turn of the century was a really big change for me because it opened up lots and lots of opportunities both to build models and apply them. And that's what led to some of the commercial work, and on and on and on.

It’s been just really, really fun to watch my horizons expand as the world's horizons expand. A lot of young faculty and sometimes undergraduates want to kind of, you know, almost replicate what I've done. But—and I'm not saying this with any kind of false humility, I really mean it—it was right place, right time. I just happened to be hitting the ground here at the time when all these data sources were just opening up. There weren't that many people doing it, and I'm just lucky that I was able to ride that wave and continue to do so.

It's much harder today because there are many more smart people wanting to do these kinds of things. So it's just nice to get in early.

16:47 Nora

Nice. So you got all this flood of data that started coming in. Did the methodologies that you were using to study customer behavior change as the data started?

16:58 Dr. Fader

Yes, indeed. A lot of the work that I did at the beginning—again, I was working with a lot of packaged goods companies, you know, the Nestlés and the Kellogg’s and all them—to basically help them figure out how to sell more toothpaste or whatever. We’ve got this new product coming out, so how can we, you know, how can we use our marketing dollars more effectively? It was very product-focused. 

Around the turn of the century, for a variety of reasons, things became a bit more customer-focused. So instead of figuring out how to sell more of this product, it became more about how to sell more stuff to this person? And therefore, what kind of stuff should we be developing, and how should we sell it?

It really was kind of a 90-degree pivot. Same methodologies to a large extent, but a very different set of applications. And I made that pivot less because of changes in the world and more because academic reviewers reading my papers were saying, “We’re getting bored with a lot of the new product forecasting stuff.” It was like, “Come on, Fader, you’ve got to come up with something new.” So I made this slight pivot to use similar kinds of methods but do customer-focused work instead. And that really did open up lots of possibilities.

A lot of the work I did in the early 2000s was the idea of customer lifetime value. What is going to be the future profitability of each customer? And boy, oh boy, does that open up all kinds of possibilities—both from a math and modeling side as well as from a managerial practice side.

I really started pushing hard to get companies to take a more customer-centric view instead of a product-centric one because it let them leverage the data better. It was a little self-serving because that’s the stuff I found more interesting. And again, right place, right time. A lot of companies just happened to be looking in that direction, realizing that we can only go so far just pushing products—“buy more of our stuff.”

The idea of figuring out differences across customers and deciding which message we should send to which customer at which time—a lot of things we take for granted today—were just starting. So I was, again, right there, basically helping companies do that more effectively, both in terms of the models they were using and the tactical things they would do to leverage the models. And that’s when it started becoming irresistible for me to not only continue the academic work but also to go out there and start companies. To actually hold companies' hands and help them do this stuff. It was a way to showcase the research, a way to get data sets for new research, and just, you know, to change the world. To have some genuine impact instead of just being satisfied with having another line on the CV. I really want the research to go out there and be understood, appreciated, and applied.

19:52 Nora

Yeah. Could you actually talk a little bit more about your work in the industry—just venturing into founding startups as an academic? How’s that?

20:00 Dr. Fader

Yeah. So I've always been interested in it. Unlike a lot of academics who find that stuff dirty and disgusting, I've always wanted to have some kind of impact. A lot of it was in packaged goods, the music industry, and online companies. I’d go to companies and say, “Hey, I have these models. They're cool. They work really well. You should try them out. And oh, by the way, here's some spreadsheets, R code, technical notes, and videos. Go, go, go—use this stuff. It’s good for you.”

And they would ignore me. They would say, "You're just an academic. You don't really understand our company. Our company is different."

So I had to take matters into my own hands. I took this as a personal challenge. In the late 2000s, I started doing two things. First, I started writing a lot of lightweight books. In addition to all the academic work—which is, you know, heavy-duty and technical—I started, I hate to say "dumb it down," but focusing on the applications and motivations behind the models. I wrote all these books on customer centricity, focusing on the right customers for strategic advantage.

And it's stuff that's, you know, mildly embarrassing because it's not what I really do for a living. It's not what I'd want people to think about me. But if it can get people interested in the research and maybe lean into the models a bit, it's well worth it. So I started writing some of the books, and that was good. That was great. But it wasn’t enough. I could lead the horse to water, but I needed to find a way to shove his head into it or something like that. And that's when I started commercializing the models themselves. Instead of going to my students, who found this work very compelling, and they'd say, "Hey, can we use this in my startup?" I’d say, "Yeah, go for it, sure." But I realized I had to take matters into my own hands.

So, working with a bunch of students, we co-founded the first company, Zodiac. All the co-founders were current or former Wharton undergrads. And it was great fun to bring the models to life at full commercial scale, working with a wide variety of companies and showing how all these things work and the implications of them. And of course, as you mentioned earlier, Nike bought that company. That was an amazing, amazing outcome. Not only is it just nice to sell your company to a behemoth like that, but a lot of folks who were skeptical—a lot of folks who had dismissed me as a mere academic—all of a sudden were saying, "Whoa, I guess this stuff is real."

So it created a whole lot of interest. A lot of other companies started lining up, saying, "Hey, work with us next." It was very gratifying to see companies on their own—instead of me pushing them—saying, "Hey, can you tell me more about that stuff?" I have to admit, it’s very addictive to get that kind of gratification from outsiders.

But the good news is, it’s a great source of data, a great source of questions, and a great source of brilliant students who start coming up to me saying, "Hey, can I work with you?" It’s been really, really nice as a way to fuel some of the ongoing academic work while continuing to push for advances in practice.

23:11 Nora

Nice. So it was acquired by Nike. Could you share some of the success stories that you've had with using your models and predicting customer behavior, and how that changed outcomes for some companies?

23:22 Dr. Fader

Sure. A lot of it, in the old days with Zodiac and Nike, was just tactical marketing stuff. Like I said before, helping companies acquire customers more effectively. For a lot of companies, the whole thing is: can we acquire as many customers as we can as cheaply as possible?

And I’m saying, no, that’s the wrong way to go. Let’s acquire better customers, even if it’s going to cost us twice as much to acquire them. If they’re 100 times more valuable, it’s worth it. So a lot of it was just tactical—helping companies figure out how to acquire, retain, and develop customers. How to make customers a little bit more valuable, figuring out what to sell them, what messages to use, who to connect them with, and so on.

That was all great and good. But after Nike bought the company, they put a really tight non-compete on us. All these companies were lining up saying, "Hey, work with us next," but we couldn’t—because all of that belonged to Nike for the next three years.

But they gave us one out, one avenue we could pursue. We couldn’t do individual-level analysis—like, what is this customer worth, or how long are they going to stay? But we could add it up and make statements at the company level.

So after selling Zodiac to Nike, my co-founder, a former Penn Wharton undergrad and M&T student named Dan McCarthy—who did the finance thing for a while, then came back and got his PhD with me—he and I co-founded our new company, Theta, to bring the models to life.

Same models, but we focused on financial applications. So, like in the private equity world, when some big private equity firm is thinking of going out and buying some, I don’t know, digitally native men’s underwear company, and they want to know what they should spend on it. And they have all kinds of rules of thumb that they use to figure out what they should pay for a company. And we’re saying, “we can do that better,” because we can tell you 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.

This idea of customer-based corporate valuation—taking the very same models but using them for financial applications instead of, or in addition to, marketing applications—and man, oh man, does that have impact.

So instead of just saying, "Hey, we can acquire a slightly better customer," which is nice (and I still love doing that), but we can now help companies buy and sell other companies. And the models start having an impact in the hundreds of millions of dollars. And even though I’m not a finance guy—and never will be—when you see that kind of impact, it feels really good.

Then, after the private equity firm buys the digitally native men’s underwear company, then we can work with the company to help them do the marketing tactical stuff. So we kind of win in both ways. We have this broader set of applications, but it also gives us greater credibility and greater buy-in to come in and do this stuff.

So it's been—and again, for me, it’s been—a great education to learn about the world of finance and investing. It’s stuff that I knew a tiny bit about as a Wharton professor, but not nearly as much as I should. It’s been fun to learn about that and to figure out—here we go again—how my models can apply there, and how some of their models, frameworks, and paradigms can be brought into my world too.

26:52 Nora

Wow. And you're talking about how math and statistics can be really powerful tools, obviously. But just me personally speaking as a psychology student, I'm familiar with consumer psychology and consumer neuroscience, and I think you touched on that a little bit before. So I was wondering if there’s any crossover with those fields?

27:10 Dr. Fader

Wonderful, wonderful crossover. Yeah, that's why the field of marketing is so cool. Like I said before, half of us are quant, and half of us are psychologists. And the best, best papers are when we work together.

Far and away, the most impactful paper I ever wrote was one where we took a very interesting extension of prospect theory from Tversky and Kahneman—a multi-attribute generalization of prospect theory—and figured out how to bring it to life using, you know, orange juice data. It was just amazing to demonstrate the validity of this theory, talk about its applications, and then start thinking about lots of other applications besides orange juice.

Then, I turned around and worked with someone in our operations department to figure out how to use a similar kind of model to sell seats for a symphony orchestra. So, you know, how are people trading off things like price, and location, and the attractiveness of the event? And how do we figure out what price we should be charging for which ticket, and how those prices should change over time based on demand patterns?

It couldn’t be more different in terms of the application than the original domain where we did some of this work. But it’s just super fun. And then that carries over. Then I started doing a lot of work with Major League Baseball about pricing, seating, and stuff like that.

So yeah, it all started from literally a water cooler conversation with a psychologically-oriented colleague, and then it turned into this bizarre variety of other extensions and applications. And you know, I’m living like a Forrest Gump life—one day I’m testifying for a digital music start-up, and the next day I’m sitting in the commissioner’s seats at a World Series game.

It’s just really fun to get involved in these applications, to pinch myself and realize, this isn’t real, OK, this is all just make-believe. I don’t belong here. But I’m having a hell of a good time while I am.

But of course, more importantly is the day job—teaching the courses, working with students, doing the research, being a good colleague. I mean, that’s the most important thing. But it’s nice to enjoy some of this outside stuff as well.

29:37 Nora

Wow, nice. Yeah. I work with another marketing professor, Dave Reibstein. I would say he's a little bit more psychology-oriented. But yeah, he also talks about how he's flying to Vietnam, then going to Croatia, and all these different adventures he takes.

29:51 Dr. Fader

Yeah, Dave’s a really good example of someone who straddles that line between the psych stuff and the quant stuff. And again, he’s not the only one. That’s one of the things I mentioned that makes the field of marketing pretty unique. It really is this wonderful mash-up where we can take the best stuff from statistics, and psychology, and economics, and computer science, and anthropology, and cherry-pick different models and frameworks to blend them into a kind of witch’s brew of new ideas that wouldn’t happen in any one of those disciplines.

And again, it’s just a fun place to sit around with our colleagues, compare notes about our training, and all that. We’ve been very, very lucky, and I have to say that the marketing department here at Wharton is especially good at this. Some other departments tend to over-specialize—the quant people do their thing, and the psych people do their thing, and they don’t talk much together.

Here, we love talking to each other. We love coming up with these crossover applications, and that’s one of the things that really makes this department unique. It makes it easy to hire world-class people, it makes it easy for our junior people to get tenure here because they’re doing such fascinating work, and it gets more companies and other institutions to pay attention to what we do.

31:17 Nora

Wow. So could you talk a little bit more about how marketing as a field has evolved over time and how you see it going into the future?

31:24 Dr. Fader

It’s been amazing. Going back to my own personal story, a lot of the things we’re doing in marketing today would have been unthinkable back around 1980 and earlier than that. The field evolves as both data practices evolve, as well as our understanding of consumer behavior, and as different policy issues arise. So just a lot of the privacy stuff going on these days, like over in Europe, the GDPR that limits how much data companies can collect and what they can do with it. On one hand, that’s a big challenge—like, “Oh my gosh, we can’t get the data we used to get.” But on the other hand, it’s a wonderful academic opportunity. How do we build our models? How do we come up with best practices in a world with much more limited data?

So, whether it’s quant methodologies—which is the main thing I do—or a variety of tactical marketing and finance types of things, as well as both developing government policies and figuring out how companies should respond to them, the field just keeps getting broader and more interesting. The people coming into it often bring domain expertise from other areas. We learn from them, and they learn from us.

It’s nice to be in a very dynamic field. It keeps you young, it keeps things interesting, and it keeps the students coming back for more.

32:58 Nora

Wow. And within quantitative marketing, would you say there have been some notable milestones within the past decades?

33:07 Dr. Fader

Yeah, it’s good news and bad news. So, for me, I’m a probability modeler, right? I just want to say, “What’s the probability they’re going to do this or that?” And then there’s all the machine learning stuff going on. A lot of those techniques are really, really powerful, but there’s still a lot of room for my kind of work, and some really good room to bring both kinds of modeling together.

Unfortunately, I have to spend a lot of my life convincing people why machine learning—and all the different techniques under very that broad umbrella—aren’t necessarily the be-all and end-all. There are other kinds of techniques, like some of these probability models, that actually offer a lot of value and sometimes can actually outperform machine learning techniques.

So, it’s a job I take very seriously: both to broaden people’s skill sets and give them an understanding of which tool to use under which kind of circumstance. That’s a really big part of what I’m doing—both in the classes, in my research, and commercially, as I figure out how to combine these different methods together to come up with something more powerful than either kind by itself.

34:21 Nora

Wow. So if you were someone taking a marketing class with you, you’d be learning all these different methodologies within marketing—not just machine learning, but also probabilistic modeling?

34:32 Dr. Fader

That’s right. And this course that I teach—it’s just math, math, math, math. We’re solving integrals from day one. Even though it’s a marketing course, there are no group projects, no guest lectures. It’s just math, math, math, math, math. And then a big, nasty final exam at the end. It’s old-school.

And it’s just remarkable how people keep showing up for it because they see some of the commercial success I’ve had, and they think, “Oh, we’re going to learn how to build a business.” No, you’re not. You’re going to learn how to solve a certain kind of integral. But, by the way, you can build a business around that integral.

I’m a math guy, and it’s great to find kids who, for whatever reason, loved math in high school, but then they come to college and it kind of loses some of its allure. I want to be the Pied Piper that leads them in the right direction—that shows them that love for math was actually well-founded. They just have to go in this other direction instead and think about maybe some of the practical problems we focus on.

It’s been so nice to find kids who are really serious about math. Some of them wanted nothing to do with marketing or Wharton at all. I remember one of my former undergraduates who said he never wanted to set foot in Huntsman Hall—and, of course, now he’s a marketing professor. Because they start to see that actually, we can make math beautiful, and practical, and interesting.

I really enjoy doing that. I’m an old full professor here at Wharton. We only need to teach two courses a year, and I teach six. There’s no reason for it outside the fact that I love doing it. Undergraduates, MBAs, executive MBAs—anyone who’s crazy enough to listen, do the math, and think about the applications—step on up. I’m really happy to talk to them.

36:38 Nora

Nice. So if there are some Penn students out there who are interested in working with you, would you say it’s helpful to have that math background and an interest in math?

36:48 Dr. Fader

It’s very interesting—first, for this course that I teach, I have no prereqs [prerequisites] for it, even though it’s a high-level course, what is it, 476? And I make that very clear. I don’t really care what kinds of courses you’ve taken. I mean, if you’ve taken a few math and probability or statistics courses, that’s great. I wouldn’t hold it against you. But all I want is aptitude and motivation. I just want smart kids who are willing to roll up their sleeves and do this stuff.

I remember one of the earliest TAs that I had, Preeti Karthikeyan—she was a double major in English and History. Why she would take this course, I can’t imagine. She had never opened Excel before and took great pride in never having done anything like that. And she ended up being a superstar TA for me.

So it’s really great to find not just those kids who know they love math, but also those who, for whatever reason, didn’t think math was for them—maybe because of some trauma they had in high school with it, I don’t know. And to show them, you know what? If you’re smart and resourceful, I’ve got a place for you.

So yeah, it’s good if you’ve taken some of those high-level math courses, but it’s sometimes even more rewarding when I can find people who didn’t think they were a math person and show the world that they are.

38:14 Nora

So you’ve talked about some Penn students that you’ve worked with. Do you have any last words of advice for Penn students?

38:21 Dr. Fader

Sure. First of all, you might be good at things that you’re not aware of—a lot like math.

I really want all students to take full advantage of this wonderful university and don’t over-specialize. You’re going to major in something, fine. But I want you to take courses as broadly as possible because one day, you’re going to get hit by that lightning bolt that says, this is what you were meant to do. For me, it was apparently being a marketing professor.

Unless you open yourself up to it—by talking to a variety of people in a variety of schools, exploring a variety of different application areas—there’s less of a chance that you’re going to get hit by the right lightning bolt.

So yeah, take full advantage of this wonderful university. Get to know your professors, too. Too often, we think about the person in front of the classroom and it’s just all about what’s going to be on the test. Maybe you go to lunch with the person, and all you talk about is the course.

But get to know the faculty as people. Find out about their quirky hobbies, their backgrounds. I enjoy that—getting to meet other colleagues as well as students.

I really want students to get involved with CURF and see what opportunities are out there. Think about careers that mom and dad might not have planted in your brain when you were a junior or senior. I’m really, really happy to talk to students to help them find their interests, even if it’s not my stuff, but just to get them thinking more broadly. To lean in a little bit more, whether it’s doing research or just taking more academically rigorous courses.

It’s just great to see those kinds of things happen.

40:04 Nora

Yeah, that’s awesome. Thank you so much for taking your time to be here with us today.

40:08 Dr. Fader

Truly my pleasure.