--:-- Host: Hello everyone and welcome back to the show. Today we have an absolutely tremendous episode for you. It is a conversation with Dave and Helen Edwards, authors of *Make Better Decisions*, a tremendous new book with the subtitle of *How to Improve Your Decision Making in the Digital Age*.
--:-- This is really a fantastic conversation. It's one of those conversations that is why I love doing these podcasts. You get to meet new people that you didn't know who are doing awesome things with great ideas. We talk a lot about how to make great decisions, how to integrate those decisions in the massive amount of data that we have. What is the value of data? When should we use data? When should we go with intuition and instinct as leaders?
--:-- This is a fantastic conversation. I took quite literally two and a half, three, three full pages of notes during this conversation. I could have talked to these guys all day. I have the book, I'm reading the book, it's wonderful. It is very much something that is worth picking up. You can get *Make Better Decisions* on Amazon or anywhere that books are sold. You can always go to the show notes for the page and find the book link there if you want. Wherever you consume books, you can find this book. I highly recommend it. I really like it. I think you're going to know exactly what I'm talking about after you have a chance to listen to Dave and Helen and their thoughts on how to make better decisions. It's a tremendous conversation.
--:-- Before we get there, make sure that you are subscribed to Finding Peak. Go to FindingPeak.com. It is my new Substack. Free content coming out every week around peak performance in business, in life, and in insurance, specifically tailored to us, the insurance industry. We do a wide range of topics, everything from personal development to leadership development, development in business, our relationships, and also deep dives into marketing, into lead generation, into digital sales, into what we're doing at Rogue Risk to be a human-optimized digital agency. Very much the model that I believe is the future of the insurance industry, the future of the independent agency. If you want to learn how we're doing it, go to FindingPeak.com, subscribe, get the emails, and if you want the deep dives, you can pay for that, which is like $7 a month.
--:-- I appreciate you guys for listening to this show. As always, this is a labor of love, and I just love that you guys give me your time, so I appreciate you. With all that said, it is time to get on to our conversation with Dave and Helen Edwards, co-founders of Saunders Studios and the authors of *Make Better Decisions*.
--:-- [Transition to interview]
--:-- Host: Awesome. Well, I'm excited to talk to you guys.
--:-- Dave Edwards: Thank you.
--:-- Helen Edwards: Same here.
--:-- Host: I went through and looked at a lot of what you're doing. I think that it's incredibly relevant to the audience who listens to this show, which is primarily insurance professionals from up and down the spectrum. Our audience is individuals from one-person startup agencies in small town, wherever America, to executives at the highest level in corporations in Hartford, Des Moines, Columbus, and all the places where insurance companies operate, primarily in the US. So just so you know who we're talking to today.
--:-- Normally I like to get right into the show. I'd love if you guys maybe start with your origin story. Obviously, I've done some background, but I'd love to hear it. Every good superhero duo has an origin story. Maybe we start there and dig into some of the stuff that I think is incredibly relevant to what's happening.
--:-- Dave Edwards: Okay. Rock and roll. Sounds good. Well, thanks for having us. I guess we've been working together for more than a decade. I've lost track of the number of years. We've started multiple companies. Some have worked out, some haven't, as well as the others. Saunders Studio has been around since 2019-ish, and it continues work that we've done for several years.
--:-- We started off working really closely around thinking about how humans and machines come together and what's happening to us as individuals as we are digitized, as our behavior is being monitored, as our communications are being managed, as the algorithms are making decisions for us and pointing things out in the world, as we've been put into finer and finer grain buckets and being ultra-personalized, but in a way that we can't interrogate and understand because we can't see it.
--:-- We sort of started Saunders Studio with the mission of really wanting to open people's minds to the richness of humans while we're in this digital age. It's not us being supplanted. It is actually, where's the beauty? Where are the wonderful parts of being human? And how do we help people understand that?
--:-- Helen Edwards: When we kind of got into this, the zeitgeist, if you like, was very much an either-or. It's either machines or humans, and the machines are going to rule us all. The more we looked into it, the more we did the research, the more we talked to people, the less we were convinced of that story. This is very much a, how do we do both?
--:-- Dave Edwards: We spent time with organizations that sit there and say, "We've spent all of this money on these huge data projects and putting all kinds of AI into organizations to make more predictive analytics. And it's not really working, or people aren't really using it, or they feel like decisions are harder than they would have been before with all of this data. What do we do about this?"
--:-- That was the genesis of the book, *Make Better Decisions*. It was helping people really understand the core of who we are as humans and who we are as decision makers, as individuals, who we are as team decision makers, and how we think about making decisions with data and with AI. That book accompanies workshops that we do with large organizations. We also have started working in complex problem solving, which is an interesting area of thinking about complexity and how you think about solving complex problems in a way that is quite different from simple or complicated problems. There's lots to unpack there.
--:-- Host: That makes my job very easy. I'll give you some context to some of the issues that we're facing specifically in the insurance industry, and then I think it's going to be highly relevant to what you guys do. I think we'll have an awesome conversation here.
--:-- I actually own an independent insurance agency called Rogue Risk. One of the very first things that I wrote down was the term "human optimized." What I meant by that was not necessarily all the way to an AI/ML situation. However, what I realized throughout my career, having done this for 17 years and spending a lot of time on the traditional side, is that the all-human version of our business was dying. There are plenty of boomers that are holding on for dear life to the paper, file cabinet, very human, all-human version of this business, and they've been highly successful in that method. But we are rapidly changing into an ecosystem that most industries have already moved into, which is this mixed up, mashed up situation. What is the value of data? What do we use? What data actually allows us to have better outcomes? How do we capture it? Who owns it? There are all these crazy decisions happening.
--:-- The premise of my agency was that there are moments that add value and there are moments that don't. I want the humans only spending time, or spending the most time possible, in the moments that add value and have systems, processes, use data, feedbacks. Eventually I think we get to a place where we're using some form of AI. I've been playing around with OpenAI a lot lately to handle those processes that don't add value, the humans don't add value to. You waste a lot of time, energy, resources, brain cycles throughout the day doing stuff as a straight human that doesn't matter.
--:-- So where do you mash those up? What value actually is value? These are enormous questions in our industry. I think something like 80% of the COBOL programmers left in the world are employed in the insurance industry in the United States. It's this snap forward of technology, and now we're having these types of conversations. I don't think anybody has an answer and no one is doing it well.
--:-- Unpacking what you said, one place that I'd like to start, just because it's an enormous buzz term in most industries, but certainly in ours, is data. The two questions that I wrote down were: Can we have too much value? Can we have too much data? What does that look like? Does data even have value? I've had this conversation multiple times on this podcast. Is it the data that has value, or is it what you bring out of the data? What does all that mean? Let's start with the actual nuts and bolts and try to get to how we use it to make better decisions as we go.
--:-- Helen Edwards: Data definitely has value once people understand it, or machines understand it. Let's start with that. Why do we even want AI in the first place? It's because humans think in one, two, three, four dimensions, and machines can think in unlimited dimensions. This ability for machines to take data that in some cases is really quite alien or seemingly inhuman, collected below our conscious, outside of our conscious recognition—eye tracking, mouse clicks, and things like that—the machines can find patterns in that at enormous dimensions. There's really no practical limit other than compute power and cost.
--:-- The problem with that is that eventually, for most situations, a human has to be able to justify how they use that prediction from a machine. If the prediction from the machine is decoupled from the human level decision making, which is what you'd expect in most human facing products, then we need to have accountability and responsibility. We need some sort of justification, which generally means some kind of causality, not just a correlation, which is what the machines are good at.
--:-- Humans come in because it's only us that can really put the causation in and put the justification in and say, "Yeah, the machine says this, but we're going to do this, or we're going to do this anyway, or we're going to follow what the machine says, and this is why." That level of responsibility and accountability sits only with humans for now. The danger is that we fail to recognize that. That's really what the AI ethicists work on—how do you stitch together hidden bias in data sets and make the right decision on top of that.
--:-- Dave Edwards: [Jumping in] I would add that what matters with humans is that data can be utterly overwhelming. That high dimensional space is completely unintuitive. It's like trying to apply quantum theory to something. People make mistakes on that basis and become overwhelmed. A lot of the dichotomy we hear around whether data is of value, is it too big, is it this or that—it's actually more about how humans tackle overwhelming amounts of data. That's really what our book is about—ways to not be overwhelmed and to recognize when human cognitive biases work against you in your judgment and decision making.
--:-- The distinction for me is that data has value, or at least hypothetically has value. There can obviously be data that has zero value. But the challenge is that in order for that value to translate to us, we humans have to know what it means. We generally communicate things about what one thing means when we're communicating with each other by telling stories. It could be a simple story. Here's the story for why this is the primary customer target. Here's the story for why this is the right insurance product for you. Here's the story of the US American dollar, which is essentially a story.
--:-- The challenge is that data doesn't tell stories. We have to tell the stories with the data. That's a gap that is misunderstood and easily overlooked because people are used to seeing these great dashboards. You log in and you look at your Tableau and it's got all kinds of colors and lines and things. Doesn't that tell you everything you need? No, because it's not telling you any sort of story over time. It's not telling you any cause and effect. It's not applying it in a form of context that we understand naturally because we've evolved to be able to communicate with each other using stories. Data is a really recent addition into this whole concept. Even when it's presented in a two-by-two, in a two-dimensional space, we don't naturally know and understand and agree upon the story that's there. We have to translate it. That's a difficult thing for a lot of people.
--:-- Host: One of my most interesting takeaways in the move from being a foot soldier in a company to being a leader was how differently the same piece of information could be interpreted by a group of people. You present a stat on a screen to a group of 10 leaders in an executive forum, and the feedback you get from the angles that everyone slices that singular piece of data up is incredible.
--:-- We recently were acquired back in April by a larger holding company. I'm now on the executive leadership team of that holding company. All the division leaders get together, there's 17 of us total. We're walking through different departments and we see a result, our variance is off here, and we ask why we think that might be happening. Without a story to the data, the "why" of something is just personal context, filters, biases, experience. It gets passed through all these different things, and what comes out the other side is like, you almost start to think, "Which one of us is the crazy one?" We're all staring at the same piece of data, yet we're seemingly seeing completely different pictures.
--:-- I think that's where I sometimes get lost in my own leadership. How much do I trust the data and how much do I trust my gut? What does that look like? One of the things I wrote down during your origin story was, where is the nuance? How do we understand nuance in a data-rich world?
--:-- Something that scares me, probably because I read too much, is that I'm a big fan of Nassim Nicholas Taleb. Right now I'm plowing through his epic *Antifragile*. He talks about black swans and the triad. I always think to myself, when I look at data and I think patterns and pattern recognition, are we creating fragility in our business? Pattern recognition essentially starts to carve out black swan events. We start to see things as how they happen on the mean or in the average, and we don't realize that there could be this massive thing coming that maybe our gut as humans and experience could possibly see. Not always, but a pattern-recognized data set that's pushing everything to means and averages and giving you variances tends to carve that off. Is that a concern or something we have to negotiate?
--:-- Helen Edwards: There are a couple of things that you raised there. First, I'll go back to the very first thing you pulled up, which is analysis versus gut feeling. That's where we start our leadership workshops with exactly that question. We start it from the perspective that modern neuroscience is telling us that all decisions are emotional, and we kind of unpack that. We unpack what heart versus head means these days. Then we look at fast and slow thinking and how to trigger better ways of thinking.
--:-- What's really going on when it comes to finding that nuance is quite complex. You've got a mix of a bias for machine learning or a bias for automation, taking what a machine says and giving higher weight to that recommendation than you would even in the face of evidence to the contrary. The classic example is people that follow Google Maps into a lake. These things happen all the time with data because you put a good dashboard in front of someone and all of a sudden they forget to ask the good question. How do you step in and how do you intervene and know when you should ask the right questions, and what are those questions?
--:-- This is a very human process. Sitting around that table, presenting people with one data point that they see differently—we sort of give people a bit of a release from that because it's quite anxiety provoking. The promise of the analytics movement has bled into how we think about each other. The promise of the analytics movement is that there's one single optimizable answer that can be found best by a machine, not a human. We forget that all difficult decisions, by definition, are difficult because people have different perspectives.
--:-- So why do we have different perspectives? Our cause and effect reasoning causes us to think in quite noisy ways. This is recent work by Daniel Kahneman and Sibony. We have this noisy, undesirable variability in our thinking. That variability can be desirable—it's called creativity. We all have a different perspective. We'll show people perception illusions, perception pictures, and ask them what they see. Everyone sees something different. It's quite predictable that everyone sees something different. So we shouldn't be surprised that everyone sees something different in the same data.
--:-- The question then becomes, what do you do about that? What you do about that is, firstly, embrace that diversity. That diversity in thinking is what's going to get you through a complex problem. There are lots of techniques for optimizing and maximizing the human part of that. Some of it is that people do need to have what we call "minimum viable math." Especially in something like insurance, you should be sitting there knowing what a mean is, knowing what a standard deviation is. You'd be surprised, unfortunately. That's why we teach minimum viable math—to give everyone the same common language. If you're using machine learning or any kind of predictive analytics, you really need to understand what a false positive is, what a false negative is, and how optimizing for different things in different cohorts can give you unintended outcomes and affect overall profitability.
--:-- The final thing I want to touch on is what you were talking about with the black swan. This is a new product for us, moving from decisions to complexity. A lot of our traditional tools, whether they be analysis tools or processes and decision-making structures in organizations, are just not fit for purpose when it comes to this new world of complexity. Whether it's because we are sorting by finer and finer grain cohorts in the data, or because people on the other end have so much more agency—20 years ago, you didn't really know what someone thought of you, and now social media will tell you. These things can be organized. The self-organization, decentralization of control, and emergent property that is now humanity on the internet touches all businesses. Normal statistics just flat out doesn't work. We have to turn to complexity science.
--:-- There are new heuristics and new shortcuts that we can take out of complexity research. A huge amount happened during COVID in terms of understanding epidemiological models. That math is complex, but the insurance industry is probably one of the few industries poised to adopt some of that complex math to help with decision making. Until humans really have access to some of this new science, we have to glean lessons out of it. That's how we deal with the black swan—releasing yourself from this need to have every probability calculated. It's a different way of looking at uncertainty. It's not trying to say there's a 0.01% chance of X, because we know that's just too hard for people to deal with. The probabilities we understand are 1%, 99%, 50%, 0%, and 100%. Those are the only five probabilities that humans intuitively understand. I think that came from Richard Thaler.
--:-- We try and help people think in a much more dynamic, open, complex, networked way so that you can be released from the tyranny of having to grapple with uncertainty in a way that is counterintuitive to us. It opens up the team to thinking much more dynamically, to solving problems as they come at you, to being much more agile about how you use experimentation. You see the data in a different way. It really is a totally different way of thinking.
--:-- Host: One of the things that you have in your book, which wasn't a huge topic but was a topic that I was very interested in, I want to bring it up considering my audience probably is on the lighter side with some of these topics. When we're looking at predictive analytics, those predictions are based on past experience. The concept of how do we know when to step away from the data and trust our gut—I've had that framed to me multiple times when I'm dealing with big dashboards.
--:-- Having been in business for 20 plus years, I think it's undeniable to say there are moments where you look at everything the way that it looks and you're like, "Nah, we got to go this way. Here's the answer." And you're like, "Why?" I'm not a hundred percent sure. I see this and I see this over here and I feel this, and I just can't explain a hundred percent other than I know this is a direction we at least have to try. That's a really hard call. Those calls are becoming even tougher now that we have so much data behind every decision. You struggle to justify it.
--:-- One of the things I've seen in some of the organizations that I've been in is that more data leads to more bureaucracy. People are less willing to take chances because those chances aren't necessarily backed up by the data. How do you, as a leader—my style tends to be more wrecking ball than craftsman—how do you know, or what is a good heuristic for when to take the leap away from the data and when to stick to it? I know that's not an easy question, but I know it's a question a lot of people in our industry in particular are dealing with. It feels like we should go this way, but the data is telling us to go here. How do we manage that divergence?
--:-- Helen Edwards: I think it's a terrific question. I think it's the core of where we all are right now, because what you highlighted is that the data can make us quite risk averse. We need the next data point. If so much data is available, surely the answer's there. There are a couple of things, and this is really why we wrote this book. You can't go head on into any of this. There's a subtlety. The real nuance is being able to look at it from lots of different angles. Pick up your wrecking ball and turn it around a few times.
--:-- The first one is feelings. There's no question that feelings come first. If you don't like the way that graph looks, you're going to feel it. That feeling is going to impact how you evaluate it. It's part of how you process information. We don't sample from our brains in a way that a computer does. It's a probability distribution. Depending on how we feel, we're going to take a different reaction to data. So the first thing is, how are your feelings actually influencing the way you process information?
--:-- Another nudge that I use all the time is ambiguous data. If the data is unclear, if the answer isn't in the data, then we have a natural tendency to use our intuition and our gut. As a leader, you step back and say, "Why is the answer not in the data? Is this question actually not able to be answered by the data? Or is the data not representative in a way that helps me make a decision?" So going to that next level of, do I want to use my gut because the data is not clear, or do I want to use my gut because of some other reason?
--:-- There's a lot of interesting research that founder-led organizations are able to take a lot more "step away from the data" moments. That's because the founder has more scope. They're seen as more able to take risky bets because their name's on it. There's an accountability thing. Being able to decouple what the data says is the right decision from the decision that is made by the actual human—that's okay. Data is past events. It's relying on a stable world. It's possibly biased. So are people, but there's going to be bias in there. Data is not imaginative. It has no ability to make transformational creative leaps. It can only be used in the service of those things. In the end, it's totally fine that a human makes that decision. But I think we have gotten ourselves in a little bit of a knot because of this promise of the analytics movement that the answer's in the data. It may not be.
--:-- Dave Edwards: Let me go back to what you started with around feelings. I think it's an important one, especially as you described yourself as saying your leadership style is using a wrecking ball. My question is, what's your emotional sweet spot for making a good decision? When we're highly charged, when we're highly stressed, we will lean more on intuition. We've evolved to do that. That's why we run when something is really stressful, when something is scary. Those kinds of emotions will lead us to use our intuition more. The question is, what is your emotional sweet spot that allows you to find the places where your intuition is actually reliable?
--:-- Intuition is great, by the way. It's cheap, cognitively cheap. It's generally good enough. It is all based on data, meaning the data of our own individual experience. But it is something that is quite useful. The question is, where's the emotional sweet spot that allows you to say, "I'm going to slow down and I'm going to consider this a little bit more. I'm going to think through this."
--:-- I think the next step would be to really evaluate one of the nudges in our book, which is about experience versus data. When you look at that data and you go, "I don't think so," stop and query what it is about your experience that's different from what the data is telling you. Then think about how those two might be combined. You might want to rely on one or the other, experience versus data, or combine the two of them.
--:-- For instance, we recently did some work with a big retail operation. The data about what happens in the retail stores can differ from individual experiences working in those stores. That makes sense, right? Large data, individual experience. Sometimes one is more important than the other, but sometimes you have to mesh them together in order to make a decision. You can't just blindly follow one or the other. You have to go into it and learn from the extra context of, "Well, my experience is different from the data, and here's why. Now that I know why, what do I want to do with that?"
--:-- Host: Resolving that anomaly is, I think, a really important step. And it's actually really fun to do. If your gut feeling is telling you something really different than the data, like you said, you explained the process of digging into it more, but resolving that anomaly can be extremely satisfying. It's one of those aha moments.
--:-- Dave Edwards: Oh, for example, we use a fun little case study that came out of Tim Harford's work about his experience of the London Underground, where the trains are just packed all the time, but the data collected by Transport for London suggests that the trains are empty. And he was like, "This doesn't make any sense." So he dug into the data and he explained how the measurements were taken. You have to really understand the exact moment the measurements are taken, and why, and who is looking at it.
--:-- His pithy integration of the story is that Transport for London measures the experience of the trains, whereas he measures the experience of the people. That's such a lovely insight. How do you move from measuring the experience of the trains to measuring the experience of the people?
--:-- So we have these nudges that have you really dig into it from the perspective of: where exactly is that data point taken? Why is it taken? Who's looking at it? And what kind of processing happens before you see it as a chart, a graph, or a table? As you step through that process, you realize that some of this data was taken for an entirely different reason. It's measuring a completely different experience.
--:-- Host: I love that concept of "whose experience are you measuring?" I love that.
--:-- [The conversation pauses for a host message.]
--:-- Host: So a couple things. One, they actually just discussed the concept of founders making more decisions off of judgment. I don't know if you guys listen to the All In podcast, but one of the hosts, Jason Calacanis, said there's some statistic about a certain percentage of equity. Once the founder is below that threshold, they compress down. They stop taking chances. They stop stretching. They stop breaking new boundaries. They just start running the company day-to-day operationally. Any innovation slows because at a certain point of equity, they know they can be fired.
--:-- As soon as the founder hits whatever percentage of equity where they can potentially be fired, the growth and innovation just compresses way down because now they can't step out onto a ledge and come back from it. I find that incredibly interesting because I have been fired multiple times, seemingly because I have not learned whatever that trigger point is that's broken in my brain.
--:-- To the part you asked about my leadership style being more of a wrecking ball, I think the reason I prefer that method is I like to know the actual answer. I really struggle with armchair decision-making. If I try something and I get a result, then I know what the answer is, versus sitting back and saying, "We think this is what would happen if we tried that thing, so we're not going to do it." I tend to just say, "Okay, let's do it. Let's go try that thing." If it works, then we know what the answer is. I don't know if that's for everybody or the right way because it gets you in a lot of trouble. But what I do think you get is very real, tangible data points on what actually happens and what doesn't.
--:-- Dave Edwards: You definitely want to differentiate between throwing stuff at the wall and just trying things versus a good experiment. Testing everything is a good thing. If you can write down the hypothesis, design the experiment, and go do it, that is the way to not be overwhelmed by data. It's also the way to be realistic about the outcomes you're expecting.
--:-- One of the nudges that we use an awful lot, and that people come back to us after a year saying it's become one of their favorites, is called "Calibrate Confidence." It's the idea that most people are overconfident most of the time, and that's served us well as a species. You don't go and try stuff if you don't have some degree of overconfidence. If you knew everything you were up against over the last decade, how many different decisions would you have made? Sometimes it's better not to know.
--:-- But in general, it's a good idea to have a good understanding of the state of your own knowledge. Being able to calibrate your accuracy with how well you understand something is actually a pretty good thing. So being able to put a number on your knowledge—"I'm 100% sure of that," or "I'm 90% sure," or "I'm 75% sure"—enables you to think, "Okay, I have to put a number on it." As you do that, you generate a curve in your own mind as to where you sit in your state of knowledge.
--:-- But it also allows other people to flip it around and say, "80% confident? Well, why not 100? What's up with that 20%?" It forces an explanation, and explanations are generative. You don't just blurt out something; you actually have to sit and think. Most people under-explain. The minute they have to explain, it draws out new knowledge. You generate a new understanding in yourself and in others. It's a very powerful technique.
--:-- It doesn't mean you become a risk-averse, institutionalized ops guy. It means you are more able to recognize the state of your own knowledge. None of us want to program in regrets or live a life where we deny that we regretted a decision. A bit of regret is okay. You want a few false positives. You want to do a few things that were kind of wrong but were the right call statistically, to have enough risk-taking in your life.
--:-- Starting out with at least a knowledge of your own state of knowledge is really powerful. You might back off a few things you otherwise would have plowed into, but you might not. You might actually have a better perspective on why you're doing something, even though it is risky. "I'm only 60% sure, but this is a real high-stakes call. If we win this one, we've cracked this nut." You're able to differentiate between wild, non-thought-out risks and a really calculated risk: "We're doing this because if we win it, we've won everything."
--:-- Host: I'm going to butcher this, but is it fair to say that if you're trying to make a decision, data gives you the vector, the direction you should be looking? Then your intuition, gut, experience—the accumulation of what you've had as a professional—gives you the ability to pinpoint where you actually go. Data gives you a range of a direction. If you have 360 degrees, it says, "This data shows us this is kind of where we want to be pointed." Then because I've been doing this for 10 years and I've had these seven experiences, here are the three places inside that range that we want to run tests. Your intuition is the scalpel. Is that the opposite, or is that just a crazy example?
--:-- Helen Edwards: No, it's a really good example. The neuroscientist would say it's exactly the opposite. Antonio Damasio says feelings come first. Feelings and intuition will point you to the appropriate space in the decision space to look. Then data actually allows you to really hone in on exactly where that analysis outcome is. Then you add your humanity, your sense of accountability, your risk aversion—that enables you to actually grapple with the uncertainty and make the decision. It's kind of like a data sandwich.
--:-- Host: Yes, a data sandwich. I like that.
--:-- Helen Edwards: But I'm also reflecting on what you're saying about the data about founders and their percentages of ownership and their risks. I haven't seen the study, so I can't have any reflection other than hearing what you said and thinking, "Huh, really?" My intuition is telling me to question that conclusion from that data. It's because of my lived experience of which companies have been the most innovative at different eras: GE under Jack Welch, Apple after Steve came back, Disney under Bob Iger. Those are all remarkable success stories where the leader didn't have a meaningful ownership percentage.
--:-- Does that mean my experience overrules the data? No. But it does mean that if I was presented with that and thinking about using it for some purpose, I'd want to dig deeper and question it more to understand that delta. One of the nudges we have in our book is around "Who are the humans in the data?" Understanding what the data representation is. Then there's also the question of how you are drawing the story out of it. There could be an alternate example. We talk about listing what you'd have to believe to believe the opposite. Is it about the founder's percentage, or could it be about the size of the company? Are there other alternative causes to the result?
--:-- Host: I think this is really good. I forget which one of you made the point about what is good data. I didn't put this piece of information in on purpose, but they were talking specifically about early-stage companies. You go from a founder who owns 100% to when they lose that percentage. Often, one, they just got paid, so they went from usually broke to not broke. Two, they went from no one can fire me because I'm one of three people to now I can be fired and I have something to lose—by venture capitalists who are showing up and firing the CEO.
--:-- You take that same individual. They're a co-founder or the only founder. They own the majority of the company. They can't be fired. They can also go broke tomorrow. They're growing. They bring in a big investor. They take a smaller cut. Now they have some money. They have something to lose. They have a board that can kick them out. Now all of a sudden they start to play it a little safer because you don't want to make that decision against the grain of the data where the board and your VCs can come in and say, "The data told you to do this. You did this. It didn't work. You're out."
--:-- The flip side of everyone you just named—while 100% true and incredibly innovative—were also late-stage, enormous companies. Those guys had big, huge contracts and there's a sunk cost fallacy. I would believe in the people who gave them those contracts. If I'm paying Bob Iger $10 million a year plus a $50 million bonus to run Disney, I'm going to give him some leeway in making decisions. We're paying this person.
--:-- It's incredibly interesting how that one data point—"these were early-stage companies" versus "all companies"—completely changes the reference of what that story can mean. That actually worked out. I didn't mean it to, but it worked out pretty well.
--:-- Helen Edwards: The serendipity of a good conversation.
--:-- Host: I wanted to go back to that concept with the train, the subway system. "Whose experience are we measuring?" That feels incredibly powerful to me. How do we know we're measuring the right experience for our business? What's that filter system or heuristic?
--:-- Dave Edwards: It's a good question. How do you know whether you're measuring the right experience? I'm pausing because there are so many different contextual answers. If you think about the insurance industry, you've got the perspective of the insurer and the insured, and potentially also the reinsurer. You've got lots of different layers.
--:-- Let's say you're trying to assess whether a new insurance product is successful. The first question is whether you're measuring the right question, the right data, and the right perspective. You need to be a little more in-depth in terms of what the question is. Define success more deeply. Think about what you mean by that.
--:-- When you actually get to the conclusion and say, "Yes, this has been a very successful product," go through the classic Five Whys to make sure you're digging through to the right answer. Have you actually gotten to an answer that you truly think is there? Success could be high revenue for the company. Success could be low risk for the reinsurer. Success could be customer satisfaction for the insured. There are a lot of different layers of what success might mean. That's usually where people get caught—they're not sure exactly what they're asking of the data, so they're not sure which experience to rely on.
--:-- Data is generally harder to collect for the thing you really want to know about than it is to collect the easy thing. The first answer is it depends on what your goal is. If you go down a layer from that, pretty quickly you're in a complex situation like insurance and customer service. You'll probably find some sort of paradox or dilemma. You can't have perfect customer service at the same time as keeping costs down. In insurance, there's always this background of wanting great customer service and settling claims to make everyone happy, but everybody knows they're on the call with some sort of rationing process, some sort of gotcha process.
--:-- Being able to very quickly get down to the point that you know why measuring something is hard—why is this a hard problem to solve? What's the dilemma we're constantly going back and forth on? What are the poles of the dilemma? That's an important one.
--:-- In the decision side, there's a really important concept from Danny Kahneman: we tend to substitute an easy answer for the right answer. The simplest example is the right question is, "How happy am I with my life?" The easy question is, "How do I feel right now?" That happens all the time with data and measurement. Using this nudge of "right versus easy"—what is the right measure? Write that down. What do we really want to know? Then what's the easy measure? Put them in front of you.
--:-- In this world of data gathering by machine, using a product like Cogito to capture emotional responses and put nudges into a call center, there are so many things that are easy to measure that aren't necessarily the right measures. It doesn't mean you don't do them. It just means you need to be much more consciously aware of what the proxies are, and then what the right answer is—the right thing we're trying to get versus the easy thing.
--:-- Host: There are two incredibly relevant problems you addressed. One, technically insurance agents work for carriers. All the marketing you'll see says "we work for you," but that's technically not accurate. There's a term for it: a broker. But in the United States, 90-plus percent of property casualty insurance agents are not brokers; they're agents. They work for the carrier. But to get paid by the carrier, you have to convince your client to buy a product from that carrier.
--:-- When you think about that, you have two very large stakeholders who are in some cases at odds with each other. Do you care if the carrier is happy? If the carrier is happy, you get faster response time, higher and more inclusive compensation, access to additional products, special programs, special pricing. But if the carrier is happy, that doesn't necessarily mean the clients are as happy as they could be.
--:-- If you measure straight client happiness and all that matters is the client relationship, your relationship with the carriers starts to become at odds. You've signed a contract with the carrier and are technically responsible to them, but you're at odds with them for the client. That sounds good and feels good. Everyone likes to thump their chest and say, "My clients love me." But if your carriers hate you, your business is making less money. You often can't offer as good a product set. You may not get first pass into different beta programs or specialty programs that can provide greater access or better products to your customers.
--:-- That's not even to mention your employees, how they feel, their metrics, or the vendors and referral partners you work with. The principal-agent problem in the insurance industry is fairly unique. I'm thinking through thousands of conversations I've had around this problem. Where do you focus your attention? Which relationship is more important to value?
--:-- That goes back to where you began your answer: how do you define success? If success is maximizing revenue in every way, then you probably need your carriers to be happy. If you care more about the relationship with your clients, the longevity of those relationships, and the ancillary benefits of having deep, solid foundations with your clients—which can also be profitable, but probably not maximizing profit—that's different.
--:-- That's going to be different for every agency, every individual business, and who those leaders are. In our industry, as soon as someone starts telling me, "This is the way you should do it," every bell in my being starts going off. I've been part of too many different conversations for that to be true. It seems like this work needs to be done on an individual basis.
--:-- I want to be cognizant of your time. It seems like we should be doing this work on an individual basis versus relying solely on benchmarks or frameworks passed down by a consultant. We need to work with a consultant to figure out what this is individually. Is that a fair assessment?
--:-- Helen Edwards: I think that's fair. I'd actually make it even more individual. Basically, everything is moving to more personalized. I'd hazard a guess that when you started out 20 years ago, these relationships were much more one-to-one and not a lot of machines involved. Now, what if 50% of the value of that relationship is done by machine? Inside that, there's an artificial intelligence making predictions. Sometimes decisions are coupled with that prediction because policies are set at scale across the whole client base.
--:-- But if there's agency in that—if there's variability in the way you're making your judgments and decisions—this is a much more complex system. Suddenly we're down into self-organizing, emergent properties, adaptation. The decisions you make within your own discretion and judgment are fundamentally different than you would have made 20 years ago. You have totally different access to information. Rules are different. There are either more rules or less rules, more decisions or less decisions. They're on a spectrum.
--:-- That's the real reason this is so individual and unique, and why we went to nudges. In the end, this is about personal practice. It's about getting to know what you value and being able to understand how you specifically understand context, how your imagination works, how your creativity works. You're clearly a bit of a status quo buster yourself. That wrecking ball has worked well for you.
--:-- Host: Worked really well for me until I turned 40, and then it just didn't work. I don't know what it was about that transition. You can be a kind of young upstart. A lot of us who are contrarians in our younger years—that doesn't work as you get older. People expect the gray hairs. They expect wisdom. They're not as forgiving of those behaviors. Plus, there's a lot of survivor bias. You're here. It's worked. You're not looking at the people who didn't survive when they were wrecking balls. I can tell you some names of people who just didn't survive that process.
--:-- Helen Edwards: I think this world of personalization exists because we can do it. It's quite wise to think about this as an individual decision bubbling up to your organization's decision, whatever size your organization is.
--:-- Think about other industries that have gone through similar major transitions. Looking at the financial services business and portfolio management: 20 years ago, it was all about stockbrokers making individual stock calls for their clients. You'd want to work with one of the big banks because they had the flow, the trading desk right there. Their optimization was around trading commissions versus how much money they were making for their clients. That was the tension. You could go to smaller places, but they wouldn't have the same access to market timing.
--:-- Fast forward 20 years, and it's a totally different world. A lot of the portfolios you're picking are optimized around ETFs set up in large-scale research situations. If you go to the little brokers, they can pick anything you want in the market. When you go to the big banks, they're all regulated out of central research organizations for what they're allowed to give to their clients because regulations have changed. Still the same sorts of tensions: "Am I making money for the bank? Am I making money from my client?" But the whole profile of who you are, what you do, and what you can offer has changed. I have friends who've lived through that entire timeline, being the high net worth folks at big banks. Their jobs are completely different from what they were 20 or 25 years ago. But now, do they stay there?
--:-- Host: Well, that's their own individual personal decision. And that's fine. And I would echo what Helen said. Our premise in our book and our premise around decision making is that there isn't one optimizable answer. There isn't one heuristic to follow. There isn't one process to follow. There isn't a six-step process to make a good decision.
--:-- We believe this is truly a practice, which is why we have 50 nudges that help you get better. That's why we call it *Make Better Decisions*. It's more like meditation in terms of practice and thinking about what works for you as an individual inside the sphere of people that you're making decisions with, than it is about some sort of step-by-step process that you can put in boxes and have a framework and do.
--:-- That can be really unsatisfying for people when we say it. We truly believe it. We're not having some sort of easy get-out-of-jail-free card by saying there isn't a six-step process and we haven't invented it. We actually just truly believe decision-making is way too complex to be able to have a set process. You have to think about what nudge do I want to use right now, in this situation, with this topic, with this group of people, to make my decisions just a little bit better than they would have been otherwise.
--:-- Host: Yeah, I love that. I love it. Guys, I have thoroughly enjoyed this conversation. The book is *Make Better Decisions: How to Improve Your Decision-Making in the Digital Age*. On Amazon, I'm assuming the rest of the places?
--:-- Dave Edwards: Bookshop.org.
--:-- Helen Edwards: Yep. Local bookshops, wherever you need it.
--:-- Host: Where, if people want to connect with you guys in the digital space, where link? What's your spot? Website, LinkedIn? Where do you want people to go to connect with you?
--:-- Dave Edwards: Yeah. Our website is getsonder.com and you can reach us there.
[email protected] is an easy email address. You can find us both on LinkedIn. We also have a podcast ourselves called *Artificiality*, which is a combination podcast and newsletter that we host on Substack. So you can find us there.
--:-- Host: That's great. And I'm on TikTok.
--:-- Dave Edwards: Yes, we do. We do participate in some of the other social medias.
--:-- Host: How do you like the TikTok and Instagram and Facebook?
--:-- Dave Edwards: TikTok. I set a time limit. Like, it's just... oh yeah. Any more than five minutes, I'm wasting too much time, but it's just too damn addictive.
--:-- Helen Edwards: I think the interesting thing about TikTok and Instagram for us is that we wrote this book coming out of working in a corporate setting and doing workshops, but it so quickly becomes really applicable for people in their personal lives. We got a wonderful comment on a TikTok video where someone said, "You just explained why my marriage has been in the shitter for the last five years. Thank you." That was quite an eye-opening moment, and quite encouraging, especially since we're obviously a married couple. We work together. We've done this for a long time. It's kind of nice to feel like maybe this could be applicable in people's personal lives too.
--:-- Host: That's good. That's tremendous. All the concepts we're talking about, while applied obviously to business, I'm sure there's a derivative that applies very much to how you... and I really like the fact that you position it as a practice. In my own life, I very much try to approach things as a practice versus when I was younger, I think I oftentimes was shooting for the goal. It was, did I get to this thing or did I not get to this thing? Today, hopefully, maybe it's turning 40, which I did recently, viewing all changes in our lives as practices, unless it's something very acute, tends to be a more sustainable and predictable and proactive way of getting stuff done. So I love it.
--:-- This has been absolutely wonderful conversation. I wish you guys nothing but success on the book and everything that you're doing. Obviously, highly recommend this and hope everyone will check it out who's listening. Guys, I appreciate your time and I hope you have a wonderful day.
--:-- Dave Edwards: Thank you. It's been fun.
--:-- Helen Edwards: Thanks so much.
--:-- Dave Edwards: Cheers.
--:-- [Advertisement begins]
--:-- Host: Close twice as many deals by this time next week. Sound impossible? It's not. With the One Call Close System, you'll stop chasing leads and start closing deals in one call. This is the exact method we use to close 1,200 clients in under three years during the pandemic. No fluff, no endless follow-ups, just results fast. Based in behavioral psychology and battle-tested, the One Call Close System eliminates excuses and gets the prospect saying yes more than you ever thought possible. If you're ready to stop losing opportunities and start winning, visit masteroftheclose.com. That's masteroftheclose.com. Do it today.
