The Tools Of The Behavioral Science Trade: Matt Wallaert

The way Matt sees his own work

“It is important that both you and I acknowledge the perspectives from which we come, which is I have always been an internal guy. At every company I’ve ever worked. I’ve been an internal behavioral scientist. I’ve led product. I’ve been a product executive, other kinds of things as an internal person. You would vaguely say at Microsoft that I also did some sort of consulting-ish work in the sense that the biggest clients had the opportunity to work with me on problems of their own in their business, sort of as a perk of working with Microsoft and being a big giant Microsoft customer. But that’s, I think a different relationship somewhat than being a paid consultant, having this person who gives you some free advice, I think is different than doing paid consulting.”

A metaphor for Matt’s work at Clover Health

“I often likened it to building the car. You make this sort of prototype car that’s made of balsa wood. Then you sort of do a test and you make something where you’re starting to produce an engine that’s scarred out of aluminum and kind of looks more like what it would finally actually be. You both have an idea of sort of how it’s going to finally look at an idea of how you might operationalize it. And then there’s scaling, which is really like driving the car and sort of how do you do that? And it’s interesting, actually in my time at Clover, we shifted in some ways how that happened in the beginning was very much we’ll build the, we’ll do the pilots. We always said, “We’ll run the pilots. We’ll advise on a test and we’ll be around at scale if you need us.” By the end of when I left. And this was one of the reasons that I left, it looked a lot more like I built the car, I’d start driving the car. And then as the car was driving, I’d hop out and someone else would hop in. Where my team kept growing and growing and growing because the organization sort of couldn’t take on the new initiatives that it had asked for there came this sort of weird flashpoint where it was like either Matt’s team needs to get very large and just run a bunch of programs for us, which I was not super interested in doing. Or we need to sort of find a different way of doing this.”

When to harness each kind of behavioral science expert

“Consultants are about when you have the internal resources and not the guidance for behavioral science and agencies are about, you don’t even have the internal resources or for a variety of reasons don’t want to distract your internal resources, which is also totally reasonable.”

A glance at long term advisory work

“If I look at my own career, post the publication of the book, that is primarily what people come to me for now, “I have 500 project managers. I’m convinced the way we’re doing project management is not right. I’m convinced the way that you are presenting product management is right. Can we shift these 500 people from here to there?” That’s a multi-year, internal, very long sort of thing to do.”

A bold proposition

“I know this is heretical, but not every organization has to decide that they want to be good at behavioral science. If you think that they are the developing of human solutions, the changing of behavior is core to your thing. You probably don’t want to make one of the core pillars of your business something that you outsource.”

Brooke: Hi everyone, and welcome to the podcast of The Decision Lab, a socially conscious applied research firm that uses behavioral science to improve outcomes for all of society. My name is Brooke Struck, research director at TDL, and I will be your host for the discussion. My guest today is Matt Wallaert, author of Start at the End and former chief behavioral officer at Clover Health. Good help is hard to find and in today’s episode, Matt will tell us how to find it, but the importance of getting to know our problems first. Matt, thanks for joining us.

Matt: Thanks for having me, Brooke. I appreciate it.

Brooke: So I’m just going to take a little interlude here and say, before we recorded this, we talked about the importance of not feeling too scripted. So, just wanted to get that out in the air and now let’s continue.

Matt: I don’t think that’s going to be a problem.

Brooke: Let me tee you up with the extremely unscripted thing that I’m about to say: suppose you’re a senior executive in a company, you’ve got a problem and you want to solve it. You and I had not previously agreed to discuss, is these three different situations. Then I want to ask you how we navigate our way to choosing this kind of choose your own adventure situation. You have a problem, when do you, A, hire a consultant, B, seek out an agency, or C, hire a behavioral scientist within your team or organization.

Matt: Now I vaguely remember that we had this conversation before we get into this discussion, it is important that both you and I acknowledge the perspectives from which we come, which is I have always been an internal guy. At every company I’ve ever worked, I’ve been in internal behavioral scientist. I’ve led product. I’ve been a product executive, other kinds of things as an internal person. You would vaguely say at Microsoft that I also did some sort of consulting-ish work in the sense that the biggest clients had the opportunity to work with me on problems of their own in their business, sort of as a perk of working with Microsoft and being a big giant Microsoft customer. But that’s, I think a different relationship somewhat than being a paid consultant, having this person who gives you some free advice, I think is different than doing paid consulting.

Matt: When do you hire an internal person, when do you go to an agency and when do you get a consultant?

Matt: Well, look, I wrote a book called Start at the End. So I think definitionally, it has to come from the outcome that you want. And even the internal groups I’ve seen at places like there are internal groups that look like agencies in some ways. Meaning you buy an existing executive, have a problem. I like give it to this group. And they come back with a solution that has been arrived at through behavioral focus and tested through science. But they come back to me with more or less working thing.

Matt: The way I described it to Clover Health and actually our process was sort of pilot test scale. So there’s this, “We’re going to do a bunch of research to sort of figure out the potential ways to solve this problem. And then we’re going to try a bunch of things, but those things are going to be made of chewing gum and sticks. They’re not even close to the thing that we would actually scale in operation or in final format.” They’re just trying to test the concept, they might not be how we would actually do it in real life.

Matt: I often likened it to like building the car. You make this sort of prototype car that’s made of balsa wood. Then you sort of do a test and you make something where you’re starting to produce an engine that’s carved out of aluminum and kind of looks more like what it would finally actually be. You both have an idea of sort of how it’s going to finally look at an idea of how you might operationalize it. And then there’s scaling, which is really like driving the car and sort of how do you do that?

Matt: And it’s interesting, actually in my time at Clover, we shifted in some ways how that happened. In the beginning was very much we’ll build the, we’ll do the pilots. We always said, “We’ll run the pilots. We’ll advise a test and we’ll be around at scale if you need us.” By the end of when I left, and this was one of the reasons that I left, it looked a lot more like I build the car, I’d start driving the car, and then as the car was driving, I’d hop out and someone else would hop in. Where my team kept growing and growing and growing because the organization sort of couldn’t take on the new initiatives that it had asked for. There came this sort of weird flashpoint where it was like either Matt’s team needs to get very large and just run a bunch of programs for us, which I was not super interested in doing. Or we need to sort of find a different way of doing this.

Matt: When you hire internally, I have to think, I think you have to be very conscious about what it is you actually want. In my optimal world, the thing that you want is everybody to be doing behavioral science. I am largely becoming more interested in the centralized behavioral science, not necessarily having a centralized behavioral science team. Or if you have a centralized behavioral science team, the point of that team is to shift your entire organization. So it’s a little like having somebody who’s in charge of DNI. That doesn’t mean that person does all the DNI for that company. Their job is to get the company to embed DNI as a central tenant of everything that does, sorry, DNI: diversity and inclusion. Getting everybody to be more diverse and inclusive in their work. That doesn’t mean it’s the DNI person’s job to do all of the inclusion. It’s their job to help people be more inclusive. I sort of think behavioral science the same way.

Matt: When you’re hiring internally if you want to build this as a core capability at your organization, you want everyone in your organization, you want to shift the culture of your organization towards behavioral science, that is a prime place for an internal person. I also would argue that that might be a prime place for a consultant, a long term consultant. So meaning somebody who picks up your organization, which doesn’t do behavioral science, or does it only in proto now, and then drops it off at a place where people are doing it more. So these are long relationships, two, three years, there may be benefit to having an outside opinion. So there are places like that where consultants make sense.

Matt: If we go from the outcome perspective, there is an outcome that is: I want everybody in my organization to do behavioral science, the sort of teach a man to fish kind of version. Then there’s a very different outcome. I think that is the, “I just want the fish. I don’t need to know how you made it, I don’t have the estimation that I’ll need a lot of fish or I’m willing to pay for other people to catch my fish for me.” Both of those could be true. There are very good reasons why for budgetary reasons might rather just pay for it on a per fish basis. And there are reasons where you’re like, “Well, this is the one and only time I’ll ever need to get a fish.” Those are real.

Matt: That feels like a place where agencies shine, because unlikely that an individual consultant will be able to catch you a fish, I think in behavioral science. Because at least the way I think of this, this sort of project management plus quantitative sides, you could get one person to do all those, but that’s all a lot of work. It is very hard to get one person, A, it’s hard to find somebody who has all three of those skill sets. And then B, really that’s how a lot of work on a timeline that would be relatively tight, I think.

Brooke: So if you need a small fish-

Matt: A little tiny fish, maybe, or you’re willing to loan some of your own. If you’re going to short of shadow, have your own data scientists and qualitative people report to this person for the duration of the project, you might be able to do it. Maybe that’s another factor that we need to sort of add into our model as you and I are live co-developing a model. What internal resources you have matters a lot too. If you’ve never done data science, if you’ve never done user research, trying to get a consultant to run rough shod over your people and get them to do that, that’s going to be really hard. You need an agency that can add resources. Maybe that’s the best way to say it. Consultants are about when you have the internal resources and not the guidance for behavioral science and agencies are about, you don’t even have the internal resources or for a variety of reasons don’t want to distract your internal resources, which is also totally reasonable.

Brooke: We’ve got this differentiation between catching fish, or buying a fish and learning to fish. The agency is clearly in the buy a fish camp. In the teaching to fish camp we’ve got the consulting firm, which is probably going to be good to help you adopt the types of processes to get all of these different components of the engine and nicely tuned and humming along together. And then I guess hiring a behavioral resource internally is probably more about getting the right mix of components. It’s not about tuning, it’s about ensuring you have the horsepower there available to you.

Matt: Yeah, and you want to make it sort of pervasive. And I think the key on the internal thing is really internal really does shine when you’re trying to make it pervasive inside of the organization, when you’re trying to do a full-on pivot. Which is increasingly, if I look at my own career, post the publication of the book, that is primarily what people come to me for now, “I have 500 project managers. I’m convinced the way we’re doing project management is not right. I’m convinced the way that you are presenting product management is right. Can we shift these 500 people from here to there?” That’s a multi-year, internal, very long sort of thing to do.

Brooke: How do we end up in these situations where sometimes you see this kind of decentralized approach? Like if you’ve got 500 project managers, the likelihood is that they’re kind of smattered all over the organization, doing different things and helping to manage projects a little bit here and there and everywhere. So that’s a very decentralized kind of thing in most contexts. But then we also see these types of contexts where it’s like, “Well, the organization wants to adopt a new way of working.” All of a sudden we decide we want to innovate with a capital I, and so we create this centralized innovation unit and we’re like, “Cool. Now we do innovation.” What’s the difference in terms of arriving at that endpoint, which might be common to the two, which is we want in this instance, behavioral science to be pervasive in the way that we operate as a business? What’s the difference between taking the centralized and the decentralized approaches?

Matt: I think it’s interesting. So in some ways it goes back to the catch a fish, teach everyone to fish sort of version. Although there is a reason to have a centralized unit even in the catch of fish version for the ability to create demonstration projects. So much of teaching somebody to fish is demonstrating fishing. And this is I think, where sort of saying that tension at Clover, I think wehere innovation units go wrong is where operational leaders say, “Well, I don’t need to be innovative. There are some people that do this and I will just forward my problems to them.” Medicine, I think is a really good example of this. And it’s the weirdest thing because in tech, everyone wants to think of his house as innovative. And it’s interesting in medicine and other industries, I’ve seen lots of people be like, “No, I am consciously not innovative, but I know how to do is drive the car. I don’t know, want to know anything about how the things under the engine work. I don’t want to know. I’m an operational person. I’m really good at operating efficiently. Just make me a car to drive.”

Matt: And I don’t think that works very well. And here’s why I don’t think it works very well. The nature of behavioral science is that things will shift over time. That’s a really good model when you have a whole team of people who, when the car breaks, they fix it. You are really highly skilled at driving this NASCAR, but you’ve got a whole pit crew who every little thing that goes wrong with the car they can fix. And they aren’t devoted to helping you fix all of the things.

Matt: I don’t think that that’s a practical, there’s a reason that NASCAR has a very rarefied and weird thing. The practical thing is like most of us can change our own oil or do basic things inside of our engine because we need to be able to sort of repair our own things as we go. And so I think when you get people who are like, “Oh, I only want to do operational things. And that innovation gets outsourced to this other thing.” That’s a sort of recipe for success. And so then if you think of this as like a sliding scale with your innovation unit, doing all of the innovation to your innovation unit doing only teaching about innovation and they do no actual projects. There’s this middle land where they’re doing some projects, but the point of those projects is generally teaching-focused.

Matt: I think agencies sometimes sell in this way, but I don’t believe them. I think there are agencies that are like, “Well, we’re going to do it, but we’re also going to teach as we do it.” But the reason that I am very, very circumspect about that is it doesn’t align with their profit model or operational model. It is in a consultancy’s best interest to continue to be the person who catches the fish for you. And I don’t think they are consciously bad people in that they’re lying to you. I think they just have a different set of motivations that they’re not willing to sort of accept that their motivations are such that, “Hey, it really makes sense for them to try and make sure that you come to them for fish.” And you may want to.

Matt: I think sometimes people hear me coming across as, like, judgmental about that, I’m not judgemental about that. I’ve run companies. There’s totally reasonable reasons why I, as CEO might want to just get fish from people. I don’t want to have to do all my own infrastructure. I just want to go to Microsoft and get their cloud. I don’t want to do that myself. There are lots of places where I just buy fish, I think that’s okay. You have to pick and choose what your organization is going to be really, really good at. And not every organization, I know this is heretical, but not every organization has to decide that they want to be good at behavioral science. If you think that they are the developing of human solutions, the changing of behavior is core to your thing. You probably don’t want to make one of the core pillars of your business something that you outsource.

Brooke: Outsourcing and the value propositions of this kind of various players in the ecosystem. How do you choose the right consultant? How do you vet an agency? Perhaps, especially, this is a question that I find myself tangling with often is, how do you find someone who’s good at behavioral science?

Matt: I am a deeply burned track record. Well, it depends a little bit. We have to go back to our sort of basic framework here of teach people that fish, buy a fish. The people who are good at teaching people, like I’ve never subscribed to those who can’t do, teach. Teaching is very different than doing. And some people manage to do both. But I think there are people who are very good at teaching and not particularly good doers. And there are people who are very good doers that are not particularly good at teaching. You have to be conscious of that. And then I think you have to look at track.

Matt: When I interview people from my team, the canonical question that I care the most about is tell me about a time you changed people’s behavior. I don’t really care what they come up with, I think people think they have to come in with these really formal case studies from some product they built. One of my favorite answers was from a young woman who was, “It annoyed me that my roommates, after drinking wine, wouldn’t rinse them out in the sink so then the wine like dried at the bottom and you can ever get your hand in there to clean it out. We didn’t have a bottle brush and stuff. And so it was really annoying to me. And here’s what I did to sort of get my roommates to rinse out their wine glasses.”

Matt: And I’m like, “Sounds reasonable to me.” She understood why they were doing it. She took steps to change the environment that created that behavior. And then she tested whether that was actually getting her the results she wanted. Sounds reasonable to me I think that’s a great answer. I don’t need you to have done something at scale. It’s a process I care about. So maybe step one is anchor on prior process, prior articulated process. Can they do it in a repeatable way? I think the reason for this and this piece is a place where people really get wrong is there’s a lot of phenomenology. So it’s like so and so had a big title at a big company and/or had one or two big hits. It’s not actually that hard to have one or two big hits. If you just sit around and shoot a lot of bullets at a target, you’ll eventually hit the target, that’s not marksmanship. Spraying and praying down range is not marksmanship. It doesn’t say anything about your ability to do this in a repeatable way.

Matt: What you want is, “Hey, that person can hit the target every single time or with high confidence.” Maybe they missed a few times, but with high confidence they have this long track record of hitting the target, hitting target, hitting the target, hitting the target. So beware phenomenology, beware like one or two big projects that they hang their hat on. Here’s a laundry list of behaviors that I’ve changed, and the more you can get the more reliably you can sort of get in your piece.

Matt: I do think the sector matters a little bit. The process of behavioral science is universal. Behaviors and outcomes science as a process, science is a universal process. The difference between a junior behavioral science and a senior behavioral scientist, is not in process, how many times they’ve done this and across what breadth of things. There are lots of shortcuts where you can be like, “Yeah, I reasonably assume thing A, B or C is worth piloting.”

Matt: If you sit in a room with me and a junior behavioral scientist, the difference is when we put up a list of interventions on the board, I can more reliably say, “Based on previous experience, probably A, B, C and D are good ones. And these other ones are probably less likely to work. They’re not impossible. We might want to pilot them. But these ones are probably more likely to work.” And a difference in generation. The ability to sort of say, “Here’s the thing you’ve never really thought about, but I’ve used it in another context and it might be repurposable here. I do think that sector experience can be super helpful because it allows you to sort of say, “Ah, I’ve encountered situations like this before, and here are things that might have worked before that can sort of shortcut some stuff.”

Matt: You know, you’re looking for a long track record of having done it. And you don’t need two home runs, long track record base hits, a longer, particular vertical that is of interest to you, whether that’s consumer, B2B, or public policy, or whatever. So if you think about Matt Wallaert behavioral statement, population, motivation, limitations, behavior data, I think you can get somebody who has experience along anything. So you can say here’s someone who’s has a lot of experience in this population or here’s someone who has a lot of experience with this kind of motivation, or with this kind of behavior, or these kinds of limitations. Each of those variables potentially allows a specialization. And I think that’s what allows for that shared experience that can be really helpful.

Matt: This is particularly true when it comes to culture. If somebody has a long track record of being able to change behavior in the Mexican American context in the United States, or the Latin X context in the United States, it is reasonable to assume they might know some things that would allow them to do that in a representative way. That’s why diversity on teams matter so much because you get people with specialization across population, or motivation, or behavior.

Brooke: Along that line of diversity and inclusion. I think that’s a really important and valuable point to bring up here, which is that we haven’t all had the same opportunity to develop that kind of track record. I really liked the example that you brought up earlier about the wineglass stuff. That if we broaden our perspective on what counts as a valid example, we can mitigate some of these challenges, but how do we deal with the fact that some people just haven’t had the same kick at the can and developing that track record? How do we get to a situation where the kick at the can that you’ve had is not just going to be amplified time, over time, over time? Where this kind of initial difference is an early selection criteria, which then feeds into the next, into the next, into the next, into the next. So you end up with some people who’ve got crazy long track records and everybody else who’s just looking to get out of the gate.

Matt: Maybe two different answers to that. So as individual practitioners, agencies, consultants, whatever people inside of companies, creating opportunities for others and bringing them along is key. I have a long list of behavior-changing side projects often in the DNI space. I always, consciously have collaborators. I always consciously usually try and find someone who’s young if possible a woman, if possible of color, to draft into that project so that they now have something to point to and say, “Hey, look, here’s at least something in the public, in the wild that you can see that I did.” So I think resisting the temptation only to work with the established names in the field, because you think they lend you credibility, and instead lending your credibility to others, is key. That sort of credibility, the best use of social capital is actually to give that social capital away or lend it away. That’s one piece.

Matt: The other one I think is in the recruiting piece. Behavioral science is a new enough thing where it’s uncomfortable for recruiters to have to change those bullet points. I’ve told this story before where I talked to a large insurance company and they knew they weren’t looking for me. They’re like, “We recognize you’re too senior for us. We’re not trying to hire you. We’re trying to make our first behavioral science hire. It’s getting kind of a mid-level higher in the data science organization, but focused on behavioral science.”

Matt: So I talked to him for about an hour, which is a long time. And then I said, “Okay, send me your JD and I’ll see if there’s anybody in my network.” It’s a small enough field that like lots of people send me jobs and I farm them out to people.

Brooke: By JD you don’t mean Juris Doctor, you mean job description.

Matt: No, I mean job description, sorry. Job description, thank you. “Send me the job description.” So I looked at a job description and the first bullet point was, I believe seven years of applied behavioral science experience. This was maybe five or six years ago. That’s not as crazy now as it once was, although that’s still maybe a hundred people. At the time, it was probably 10, that’s a very small circle. And so then the second bullet point was seven years of insurance experience. And I was like, “Well, okay, those are not overlapping Venn diagrams. None of those seven people that have done the thing, none of these 10 people that have seven years of behavioral science experience have seven years of insurance experience. That’s just not reasonable. And even if they did, they certainly wouldn’t be taking your mid level, $150,000 manager job at some big ass insurance company. That isn’t how it’s done.

Matt: I love people without their masks. I love what happens when people are in vulnerable spaces, because I think it’s a lovely thing where we are authentic with each other. And I have had companies come and be like, “We have no idea what to pay this person. We have no comparables. What else should we be? There’s not enough people in the industry, we’re at a big company, we have some company that does Milliman or somebody that does wage surveys or something. So we have these corpuses of data sets, but behavioral science does not occur in them. How much should people actually be getting paid?” And these are hard questions, behavioral scientists themselves I don’t think know how much each other are getting paid. And so it’s hard to know how to value those services and things.

Matt: And this is not unique to behavioral science, data science had this problem. Lots of people had this problem at the beginning of their thing. I often think data science is a useful sort of context in which to think about the growth of behavioral science. Because behavioral science is sort of like where data science was 20 years ago. High demand, relatively few practitioners, relatively little information about the practice and how you actually go about doing it and how to build those things. And I’m hopeful in some ways, because data science has gotten much better over the last 20 years. And so I’m hopeful that bodes well for us, we also will sort of figure out many of those things, but I don’t think they are figured out. And so I don’t think we know necessarily how to equitably create opportunities. What’s the Kaggle of behavioral science, where you can sort of have people reasonably demonstrate behavioral science?

Matt: Now I can tell you what I tell people. So I’m relatively anti-education. So what I mean by that is I think there are reasons for these masters and PhDs and things to occur, but I don’t necessarily think that they are the best way of deciding who’s good at this thing; often, for reasons of inclusion. Given that what I always tell people is like, “Look, you don’t have to take a behavioral science. Behavioral science is an approach, not a role. I’ve often joked that, “You could have a behavioral science CFO. You could have a CFO who approaches the role of financial management through behavior change. That is a totally reasonable thing, it is an approach, not a role.”

Matt: And so what that implies then is that “Hey, you can go to your local pizza shop and say, ‘Hey, I’m going to try and help you figure out how to get people to eat more pizza.’ Or you can say, ‘Hey, I’m going to go to my local church and get people to attend. I’m going to go to whatever my community organization is and I’m going to get people to do a behavior.’” And so finding these smaller and more informal opportunities to demonstrate that skill, I think is really key. Now that does put as with so many DNI focused things. The hobgoblin of that is that it puts the onus for effort on the disadvantaged party. Which I don’t love. And that’s why as I go back to, I think it’s incumbent on people like you and I and agencies and whatever, to go create as many of those possibilities as we humanly possibly can; to act as a clearinghouse for opportunities to create behavior change.

Brooke: Let’s assume we’ve run the gamut here. We figured out what our problem is. We want to promote a strong behavioral science culture. Internally we’ve identified that we kind of have a capacity issue, it’s not just a coordination issue of getting the right pieces put together. We’ve gone through this process of trying to figure out the overlapping Venn diagrams out there in the world to hire some behavioral scientists that doesn’t exist after all of that work is done. Now we’re at the starting gate, the person arrives on the job. How do you create an ecosystem for that person? I was going to say to succeed, but let’s start with a lower mark. In large, very bureaucratized organizations, how do you create an ecosystem for that person to survive? We’ll come to success later.

Matt: This is largely about how we should allow everyone, like creating zones of autonomy. Start at the End, is written for a very particular thing, I meant it in a very particular way, but it can be read in a variety of ways. And someone pointed out to me that it could be read as a management book. In fact, managing through behavioral statements and managing through behavior change and saying, “Okay, Brooke says to Matt, ‘You work for me. My expectation is this behavioral statement. You will get Latin X women in LA between the ages of 20 and 30, who want to go from point A to point B to download the Uber app.’”

Matt: There’s no ambiguity in your performance because I have an objective way of measuring it because I know what this behavioral statement is, I know what I’m trying to create. And then you have autonomy. So you can come to me as a resource, I’ll fund things, I can unblock things for you, I can act as a thought partner, but you have autonomy to create that thing. And so knowing, “Well, what is it you want people to do as a result of this behavioral scientists being there?” I think is key from a management perspective. So that’s how you set them up for survival.

Matt: As the behavioral scientist and the thing having done this, perhaps more than anybody else, or at least haven’t done it a lot to me, there is a relatively prescriptive thing, which is you need to, within the first six months or show get some early wins. They don’t have to be big, they have to the demonstrable things that you can use as internal examples to do the teaching. In month one, or month three or whatever I’m going to do a lunch and learn. I’m going to talk about behavioral science. And all the examples I’m going to use are going to be external because I’ve only been there three months and we haven’t managed to get something off the ground. But six months, a year in, at least half of those examples need to be internal examples, no matter how small.

Matt: And I do think by the way that it’s a snowballing effect. You start with small projects that gain social credibility with people and introduce them to how to do this thing. And then you’ve done it almost hackathon style before, right where I’ve been like, “We’re going to do a one day offsite. And in the course of that one day offsite, you’re going to go through at least the first couple of steps of the IDP.” Sorry, IDP intervention, design process. You’re going to go through a couple of steps of behavioral science in order to get you to feel like you had an early win. And then that allows you to then take on bigger projects, and bigger projects, and bigger projects, and bigger projects through that framework.

Matt: I think they don’t have to own it, but they have to have access to quantitative and qualitative resources. I am generally speaking a big fan of sort of data lake access inside of organizations and relatively horizontal access to data. But I need some sort of login where I can SQL out what I want and get it without having to go through 22 layers of things to try and get that data because it’s just inherently going to slow me down. The reverse of that, there’s a lot of talk about data lakes. There’s not a lot of talk about user communities. I’m a huge fan of the Windows Insider community or Microsoft’s attempt to get a bunch of beta users together in one place. But there has to be some community within which they can do qualitative access.

Matt: It’s interesting, I’ve been doing a lot of work with a big company on trying to use one of these, “We have 500 people doing the wrong thing and we want to get them to do the right thing.” Their problem is not actually data, it’s user access. The sales teams, it’s largely a sort of B2E sort of enterprise level thing. The sales teams won’t let them talk to anybody because they’re afraid of them screwing up the relationships. And so as a consequence, they have tons of data insights, but they can’t validate any of them because they can’t get qualitative access.

Matt: It needs to be clear what you want them to do, as with anything. They need to have access to both quantitative and qualitative research, and they need to be able to get something, an intervention live into the market. And so three of the sort of tripod, qual and the ability to launch and measure interventions. Without one of those, you can maybe make it work on two, but it’s hard. Now, it isn’t coming down to behavioral scientists to be scrappy and sort of figure out their way around it. So if the enterprise sales team won’t let you have access to customers, go outreach to people on LinkedIn and find your own way to sort of noodle your way to do that. But from an organizational perspective, that’s a work around, you really shouldn’t have to do that.

Brooke: Try to minimize the need for ninja tactics to get any of these three pillars.

Matt: I often talk to particularly young behavioral scientists who have a tendency to hit a wall and then be stuck. And I’m like, “No, one of the behavioral scientists sort of greatest skills is the ability to flow water around anything. Find a way around it.” That said, if you’re managing behavioral scientists, try not to make them do that, that isn’t a good outcome. They’re very good at it, doesn’t make it a good outcome. It doesn’t make it a desirable end state and so I think your job is to then work on that.

Brooke: Yeah. So that sounds like the biggest things are around organizational process and organizational culture too, are going to play huge roles. That sounds very kind of consistent and harmonious with what you were talking about before, in terms of judging on track record, not necessarily based on education. So if we think about the way that we educate students to understand kind of the theory behind behavioral science and the historical art that has brought it to where it is and the methodological details of how to execute studies and these kinds of things; nowhere is there a chorus on how to flow like water around everything in an organization.

Brooke: Interestingly, I’ve also seen a big disconnect in people’s content of people’s practice and the form of their practice. They will be extremely adept at helping other people to find ways to get around barriers. But when they themselves encounter barriers, they are not nearly as creative or as rigorous in kind of testing these things up. So how do we take all of this really valuable education that we’re giving people? Because despite the fact that you’ve identified it as not your selection criteria, I don’t that you also think it has no value.

Matt: It’s a good clarification. I am not in any way suggesting that I think education has no value, education obviously has massive value.

Brooke: How do we create a context to help unlock that value that sometimes is just like, kind of behind the door that needs to be broken down?

Matt: Penn’s masters, what is it, masters of behavioral and decision, MBDS? Is that right, MBDS? They hook up with behavioral scientists in the industry to do a project, the sort of typical sort of MBA like internship analogy, which I think is interesting. But again goes back to the problem that you sort of identified, which is like, if they get an organization that is well set up to do this, they come out looking very smart. And if they get an organization that’s relatively poorly set up to do this, they don’t look as smart. And that’s why I anchor a little more on the question about, how have you done this? I’d rather you tell me the place where you made it happen, even if it’s small and the local church or whatever, than the place where you tried to make it happen, where success was gated on other people’s ability to sort of help you.

Matt: It’s a little bit ironic because I said, “If you want to set somebody up for success, don’t make them do it all by themselves.” And yet one of my criteria for sort of helping understand people is their ability to do it all by themselves. And maybe that’s the thing we want to hold in mind, “I want you to be able to fill any of the roles, ideally you don’t have to. But I want you to be able to.”

Matt: And so for example, on my team at Clover, we were very conscious about what we call tee time on Fridays. And so my teams are arranged into a quant, and qual and a project manager. And on Fridays there would be an hour-long top of the tee session, which we call arms time, right? The arms of the T in which one of those specialists would teach something at a general level. So a great one that Tyler, one of the folks who worked for me did, was understanding correlation. He explained correlation mathematically, and then put up graphs to sort of help you understand, how correlated is this graph actually, can you predict your strong correlations in life? How about weak correlations in life? Height and wingspan is highly correlated. Height and weight is relatively weakly correlated or less strongly correlated.

Matt: And so they would do Arms Time and then they would go into what’s called Legs Time where all of the people within a functional area, within a specialization when deep on some topic that was multilevel, hierarchical regression or something. Which would be useless to the quals or the project managers wouldn’t really find interesting or understandable, but will help them ground their team. And so we always said on my team is, your legs help you stand, your arms help you fly. And so that combination of, I am continually and always making sure that I’m able to sort of fill any of the roles if I have to. But also furthering the expertise that I have in the thing that I know how to do, I think as a potent combination to sort of helping people come forward and move along that sort of behavioral science change.

Matt: And so I often ask people who come to me and are wanting to get better at this, to identify of those three things, “What are you strongest at, project manager, qual or quant? And then which are you weakest at?” And then we try and make sure that they have opportunities to practice the thing that they’re very, very good at, at a high level. But then also can gain skills horizontally on things where they struggle and can do those things. And so getting a project for somebody who thinks of themselves a very good project manager to do a project with me where they can be a project manager, but also prompt them to do some self-learning around SQL and user interviewing and qualitative innovations in those kind of things.

Brooke: I’m feeling like that’s a pretty good point to wrap it up. I think that we’ve nicely covered the topics that we had agreed on. The one thing that you said that really stands out to me, and I think this is especially important in this diversity and inclusion perspective is the best use of social capital is to lend it out. I really, really like that. I liked the ideas and notes of service and compassion that that brings.

Matt: I think it’s interesting because it is an investment. The theory is that I lend it out and it causes that person to have differential returns, which I then participate in. I was telling somebody about this last night, it was a young person. So I said, radio play. And they were like, “What the fuck’s a radio play?”

Matt: And I said, “Oh there used to be this radio play called The Shadow at Lamont Cranston.”

Matt: And the whole thing is that when he helped you, he would say, “Someday, someone’s going to come and show you this ring and ask you for a favor and you need to do them that favor.” And so, as a consequence, as he was saving people, he was also creating this network of favors that he then could use to save more people, and more people, and more people, and more people. And there’s this sort of exponential effect where people are sort of paying it forward, and paying it forward, and paying for it. And I think that is true.

Matt: I have spent a disproportionate amount of the last period of my life, really focusing on mentoring people at the bottom of the pyramid, of the socioeconomic pyramid. And I’ve now been at it long enough where I have a great mentee, very, very proud of who I helped her all the way through college. And then she went and got her masters in dietetics and now is a registered dietician. And now literally I have things that happen where I’m talking to a new young person and they’ll be like, “Well, I’m thinking about being a dietician.”

Matt: And I’m like, “I have someone for you to talk to who has been through this journey and I know their journey and is super happy to talk to you because they had somebody help them the whole way. And so now they’re used to helping and they’re used to that process.” And so I think that what we want is, you talked earlier about sort of how do we judge people in the space and who do we look for and how do we look for them? And I think that we need to be more forward footed about investing our resources in people, who are investing resources in people. Who have a track record of bringing people along, who have a track record of sort of helping people to find their next thing.

Matt: Dan Ariely is a highly polarizing figure for a lot of reasons, but he has spent real time helping people figure out what they want to do next and encouraging them to go do that. And that’s a behavior that we need to reward. I mean, Berry Schwartz, my undergraduate mentor, along with Andrew Ward, they did things that when it was time for me to go to grad school, we all sat down and they were like, “Okay, you have a very particular personality. You’re a first generation college student. So there are a lot of things that are a little harder for you. Let’s list out all of the advisors that we think would get you.” And so my eventual advisor, Leif Anbovin really did have contextual background things that made him a very good advisor for me, because he had these experiences that were able to connect with me.

Matt: This is a little bit, it goes back to my bias about consultants and agencies and things. Consultants don’t want more consultants because then that competes with them, they’re creating these competitors. And so I often think some of them, I’ve heard them speak and I think they say things at times that are discouraging to people because they don’t want more competitors in the space. You have to be highly educated. You have to have a PhD.

Matt: You don’t hear many of the internal people being like, “You have to have a PhD.” No, the internal people are the ones going, “Hey, let’s create more of these because I need to be able to hire team members. And I need more people that I need more teams to exist, and that’s not competitive to me.” I understand. I get it. I understand people’s struggle to be a consultant and thing and I understand why they do what they do. We can end the podcast on, man, it pisses me off! That’s always a good place to end a conversation with Matt Wallaert.

Brooke: That’s perfect. So Matt, on behalf of myself, all the TDL and all of the listeners out there, I hear there are dozens of them now.

Matt: Hey, thank you.

Brooke: Thank you very much for being with us today.

Matt: Yeah, thanks for facilitating this conversation, it was really fun. I always love podcasts, where we talk about something that I haven’t talked about before. Where I’m not just regurgitating, “How do you do behavioral science? Let’s summarize the IDP.” I’m like, “I wrote a book for that.” I do think we don’t have enough conversations about how you select people? What is it like? What are we doing to cultivate us as a field in the world? How does behavioral science go out and have a conversation in a more fulsome way? And how do we recognize the way that incentives always aligned for what’s best for the field.

Brooke: If you’d like to learn more about applied behavioral insights, you can find plenty of materials on our website, thedecisionlab.com. There you’ll also be able to find our newsletter, which features the latest and greatest developments in the field, including these podcasts as well as great public content about biases, interventions and our project work.

We want to hear from you! If you are enjoying these podcasts, please let us know. Email our editor with your comments, suggestions, recommendations, and thoughts about the discussion.