Data Radicals logo

The Data Behind Dating

Seth Stephens-Davidowitz, Author of Don’t Trust Your Gut

Seth Stephens-Davidowitz

Few have leveraged Google-generated insights more than author Seth Stephens-Davidowitz, whose analyses informed the bestselling Everybody Lies and the more recent Don’t Trust Your Gut. A former Google data scientist and a visiting lecturer at the University of Pennsylvania, Stephens-Davidowitz is currently a contributing op-ed writer for the New York Times.

Seth Stephens-Davidowitz

Seth Stephens-Davidowitz

NYT Bestselling Author; NYT Opinion Writer

Satyen Sangani

As the Co-founder and CEO of Alation, Satyen lives his passion of empowering a curious and rational world by fundamentally improving the way data consumers, creators, and stewards find, understand, and trust data. Industry insiders call him a visionary entrepreneur. Those who meet him call him warm and down-to-earth. His kids call him “Dad.”

Satyen Sangani

Satyen Sangani

CEO & Co-Founder

Alation

Satyen Sangani: (00:03) There are many ways to gather data about people, like Jane Goodall, you can observe, simply note how the world behaves, or you can survey people, asking them how they feel or what they might do. Both ways are fraught. As an observer, how do you know that you have the right sample? And what if what you see is a product of something else that you can’t? Surveys are potentially worse. People might not know their own feelings or preferences. And according to today’s guest, Seth Stephens-Davidowitz, people might even lie. For example, who’s going to vote in the next election? This is impossible to predict via surveys, because as Seth puts it, nobody wants to admit that they have no intention of voting.

Satyen Sangani: (00:51) After working as a data scientist at Google, Seth realized that Google searches are better predictors of voter turnout than a survey ever could be. He points out people search how to vote or where to vote or polling places weeks before an election. And that predicts that turnout will be high. What else does Google tell us about our national psyche? Quite a lot, it turns out. Seth joins us to discuss his 2017 book, Everybody Lies, which was a New York Times bestseller, PBS NewsHour book of the year and an Economist book of the year. His new book, Don’t Trust Your Gut, offers self-help lessons across all categories — from dating to parenthood — with tips drawn from verified data. Seth’s work has appeared in the Atlantic, the Guardian, Vox and more. And today he is a contributing op-ed writer for the New York Times. This is an interview you won’t want to miss.

Producer: (01:52) Welcome to data radicals, a show about the people who use data to see things that nobody else can. This episode features an interview with Seth Stephens-Davidowitz, author of Everybody Lies. In this episode, he and Satyen discuss Google Trends, effective online dating strategies, and what the data says about our gut instincts. This podcast is brought to you by Alation. Data citizens love Alation because it surfaces the best data queries and expertise instantly. The result: folks know how to use the most powerful data with guidance from the experts, and with Alation, you don’t have to choose between data democratization and data governance. By embedding governance guidance into workflows, Alation welcomes more people to great data fast. That means your data strategy can play both offense and defense. Learn more about Alation at A-L-A-T-I-O-N.com.


Everybody Lies

Satyen Sangani: (02:45) You wrote a book called Everyone Lies. That’s obviously a —.

Seth Stephens-Davidowitz: (02:48) Everybody Lies.

Satyen Sangani: (02:49) Everybody Lies, I apologize. And that’s obviously a — that’s a disheartening statement about the human race. So tell us more.

Seth Stephens-Davidowitz: (02:57) Well, it could have been less. I think, yeah, everyone’s like, are you lying? And I say that I’m a compulsively honest person. And people are like — but I’m a compulsively honest person, there probably are some other compulsively, honest people, but I didn’t want to call the book 97% of People Lie. It doesn’t have quite the same punch as Everybody Lies, which really grabs people’s attention, but —.

Satyen Sangani: (03:19) By the way, of course, is itself a lie, because —.

Seth Stephens-Davidowitz: (03:21) Yeah. Yeah.

Satyen Sangani: (03:22) Yeah.

Seth Stephens-Davidowitz: (03:23) I guess I’m not as compulsively, honest as I say, because I used a lie in the title of my book, but I think the book is, to be honest, dark and depressing at points, I try to make some jokes in it to lighten the mood a little bit, because it does get into some arenas where there is a lot of dishonesty, things like racism or sexuality or child abuse, or do-it-yourself abortions. And the problem with understanding these topics is that not just people lie to other people, but people even lie to anonymous surveys. So if you ask people immediately after an election, “Did you vote in the election?” More than half of people who we know didn’t vote, tell surveys “Yeah, yeah, I voted. I’m a good citizen. I exercise my democratic duty.”

Seth Stephens-Davidowitz: (04:14) So if you want to learn about, say, racism, you can’t just go and survey people and say, “Are you racist? Or “Do you not like African American people? Or “Did you not support Barack Obama because he was Black?” Or “Are you drawn to Donald Trump’s message because it’s so racially charged?” But I think we have new tools to understand this. So a lot of Everybody Lies focuses on anonymous and aggregate Google search data. You see all these insights into the human condition that you’d never otherwise see.

Seth Stephens-Davidowitz: (04:43) So racism, for example, I was shocked by how frequently people were making explicitly racist searches. People were searching things like N-word jokes. And in the time period I was looking at which — thankfully they’ve gone down over time — but when I was looking at them, people were searching it with the same frequency as “Lakers” and “migraine” and “Daily Show” and “Economist.” So it wasn’t a fringe search. We’re talking about millions of people making these searches every year. And when you actually do a map of where these searches are made, you see first a surprising map that the searches are made in parts of the country that we hadn’t historically thought of as particularly racist places, like divide the United States into two regions now based on racism, it wouldn’t be North versus South, it would be East versus West. There’s much more race in the Eastern part of the country, much lower racism Western part of the country, California, Utah, Idaho … Hawaii has the lowest racist searches.

Seth Stephens-Davidowitz: (05:39) And this map, sorry, predicts all these outcomes. So in regions of the country that made more of these searches, Barack Obama did worse than previous Democratic candidates and Donald Trump performed very well in the Republican primary in places that had high levels of racism, African Americans have worse outcomes. They die quicker in these areas. There’s a bigger Black/White wage gap. So over and over again, we are seeing that this kind of secret racism that you wouldn’t otherwise see except in Google searches, predicts outcomes.

Satyen Sangani: (06:10) Is this a case where you were looking for data on research, racism, found Google Trends, or was it the other way around? What was the chicken and the egg in this original curiosity that you had? And how did you get to this place where this became the topic of a book?


Seth Stephens-Davidowitz: (06:25) When I was a PhD student, I was just constantly writing things down on Evernote and ideas I had, and most of them were the dumbest ideas ever, and they never went anywhere. And I saw the Evernote note to myself, which was that I found out, I forget even where I heard about Google Trends, and right away I said, I bet you people make inappropriate searches into this. And that’s what’s going to be interesting in this data set. So, it was very, it was conscious that the area to focus was the areas that otherwise are hard to research because …. So if you think of Google, why Google Trends is so powerful. So I’m an economist, we’re obsessed with incentives. And Google gives you an incentive to get the information you need. And Pornhub similarly gives you an incentive to get the information you need.

Seth Stephens-Davidowitz: (07:18) So even an anonymous survey, there’s no incentive for you to tell the truth. So what do I get? Gallup calls me up and said, “Did you vote in the last election?” What do I get by saying the truth? Or Gallup calls me up — and let’s say hypothetically I was gay and was embarrassed that I was gay — and Gallup calls me up and says, “Are you gay?” What do I get by telling Gallup yes to that? Nothing. But if I search, let’s say, how to vote, where to vote, before voting in an election, I get the information I need on how to vote. And those actually predict turnout very well, searches like how to vote, where to vote? If I were a racist and really enjoyed N-word jokes, then I’d want to learn the latest N-word jokes. So there’s this incentive to tell the truth that other services are… Even surveys, even though they’re anonymous, they’re not giving people proper incentives.

Satyen Sangani: (08:15) I meditate. And one of the ideas behind meditation is, “Oh well, not all your thoughts are actually serious, and most of them are ridiculous, and you should just cut and discard half of them.” But if you just take this premise of, “Okay, well look, people think things that they don’t actually act on,” How much of this do you think permeates real activity in the real world? And is there a way to correlate that?

Seth Stephens-Davidowitz: (08:37) Evan Soltas and I did a study on anti-Muslim attitudes, and people type into Google these awful things like “kill Muslims,” “I hate Muslims.” And that’s a great example of something that seems like just a crazy thought that might enter someone’s mind and they just type into Google. And it’s like, well, do we even care about that? But then we actually found that when these searches are — they correlate with hate crimes against Muslims when more people are making these types of searches, there are more hate crimes against Muslims. So, that’s proof that even though these searches are so weird and may just be a thought, those thoughts correlate. And, of course, that doesn’t mean that 100 percent of the people who made that search go on to commit a hate crime against a Muslim, but it says that some fraction of those thoughts are turning to real life action.

Seth Stephens-Davidowitz: (09:25) And another study — really important study didn’t get enough attention — not by me, by other researchers, searches for suicide on Google they found correlate with actual suicides and the correlation is much higher than surveys asking people, “Do you have suicidal ideation?” So when people are typing, “I want to kill myself,” “suicidal help,” “suicide hotline,” whatever sad sad, sad searches, we see in the data those areas have higher levels of suicide. And again, it doesn’t mean that everybody makes that search. It may be a researcher looking at what comes up. It may be a doctor trying to understand suicidal ideas more. And it may be someone who had an instantaneous thought that immediately went away and is otherwise the happiest person on the planet and just like, was really frustrated and made that search and it doesn’t go anywhere. But in those searches, there are some people who are going to, sadly, end their lives. And that’s another area where I started doing research in that, because I was very interested in, what can we learn about the suicidal minds?

Seth Stephens-Davidowitz: (10:27) Talk about an unprecedented window. What do people search before they search for suicide? And there are all these really interesting things in the data that we wouldn’t otherwise know, to my point of how this data is revolutionizing our understanding of the human minds. We don’t always know what leads someone to that really dark place. And in this search data, you see some things that are surprising. One of the things was the frequency with which people search herpes diagnosis, particularly young people then suicide. And that’s not something we really think of as causing someone to be suicidal. But we think of a 17-year-old, an 18-year-old, the mind is so poorly developed and they get a diagnosis of an STD. They think their life is ruined and over, and now some of them sadly are drawn to thoughts of ending their lives. And how can we use this information to help people in this situation?


Who Are the Most Successful Daters on Online Dating?

Satyen Sangani: (11:25) Data can reveal difficult truths about our world. It can also give us tools to improve our lives. Seth and I discussed an issue that is top of mind for many data radicals: online dating.

Seth Stephens-Davidowitz: (11:36) Christian Rudder did this fascinating study where he said, who are the most successful daters on online dating? And he found that the people who got the most messages were just who to expect, just conventionally gorgeous people. So think of Brad Pitt or Natalie Portman, they just clean up on online dating. For people who aren’t there, it’s such a grave injustice in the world, how different the life of someone in that category is in the dating world. But then he found another category of people who did surprisingly well. And it was people who had extreme looks. So think of people who … heterosexual woman who shaved their head, or people with wild glasses, or weird facial hair, or dressed oddly, and it’s like, why did they do so well?

Seth Stephens-Davidowitz: (12:24) And the point is that in dating, there is surprising variation in what people find attractive, and really in dating you don’t want to be… You want, somewhat, some people to find you really, really attractive and just leaning into some extreme version of yourself. In my case, I think I took from that, I should just nerd it out. So I used to think … I was such a nerd, I’m like, I need to be less nerdy if I want to attract a woman because we’ve long been told that the nerd is an unattractive characteristic. And I’m like, after reading this study of Christian Rudder, I’m like, “Well, no. I just need one woman, basically, whom I’m attracted to.” And I really want a date to be really into me. And I’m going to probably have a higher chance of that if I just am who I am — which is more nerdy — than I’m never going to compete with the average person, not being a normal, non-nerdy person, but I could be the nerds’ nerd.

Seth Stephens-Davidowitz: (13:22) It also relates to Pornhub where I’d say you see in the data a surprising amount of variation in what people are attracted to. So, weight, for example. We usually think that skinny people are considered more attractive, but there are a surprising number of people who are looking for porn for overweight people and people looking for, yeah, nerd porn or whatever it is. There’s always some people who are really into it. And I think the best dating strategy, if you’re Brad Pitt or Natalie Portman, just get a normal haircut, dress very nicely, don’t rock the boat. Your situation is set, just thank the gods — the genetic gods — for giving you these gifts. But if you’re not, for those of us who are not in that category, I think the best strategy is really to appeal to some group for whom you are the Brad Pitt or Natalie Portman, and then ask a ton of people out, willing to get rejected until you get shocked by someone who it turns out is really into you.

Seth Stephens-Davidowitz: (14:26) And that basically is what turned my dating life around, to be honest. My current girlfriend, her friends asked her, “What’s your type?” And they were going around in circles: “I like muscular people, I like tall people, I like this, I like that.” My girlfriend said, “I like nerds.” That was her answer.

Satyen Sangani: (14:42) There’s a couple of things that are interesting about this. And the first is that all marketing is basically positioning. So if you have a new product in the market as an entrepreneur, obviously one right here, the first thing that you learn is actually what you want to go for is your initial niche market, because being averaged basically just sticks out in nobody’s mind, and nobody actually really cares. So on some level, dating’s a market. So why wouldn’t that be true there? But in dating, I guess there’s another thing, which is that you could advertise this feature — whatever the feature would happen to be — but you might be selecting into an audience that you’re not very interested in. So I guess the presumption, if you’re going to lean into the feature, is that it’s going to actually help you select into the thing that you want to attract.

Seth Stephens-Davidowitz: (15:29) I think that’s a great point. That’s —.

Satyen Sangani: (15:29) Except that it worked for you.

Seth Stephens-Davidowitz: (15:32) But I think this advice is almost universal in that everybody probably has a version of themselves that is more extreme, that they could be leaning in more to, that they don’t lean into for fear of being rejected. And I think the truth is you are going to be rejected a lot, but that’s also your path to a great life, similarly to in business, just as you’re saying in entrepreneurship, I think everybody wants to do the average thing that everybody else is doing and not be too weird and just put a few buzzwords together, instead of doing something that’s a little weird and quirky where some people are — you’re going to go to a party and everyone’s going to be rolling their eyes like this person’s crazy — but then you’re going to have some obsessive fans.


Data-Driven Approach to Self-Improvement

Satyen Sangani: (16:15) Data leaders push their business to embrace data-driven decision making. But how do we fare in our personal lives? Seth’s newest book applies a data-driven approach to self-improvement.

Seth Stephens-Davidowitz: (16:25) My new book is called Don’t Trust Your Gut. When I was writing the book, my basic approach to the book was I just thought it was interesting that there’s all this new data in all these huge areas of life, because of the explosion of data. And that now, I gave examples in dating already, and they’re studies that I think are way more convincing than we’ve ever had about happiness and about entrepreneurship and about parenting and about making money. People have studied de-identified anonymous tax records for the first time and really gotten insights into who’s rich, who makes it as an entrepreneur, big data studies of artists of who actually succeeds, which are telling us new things. The basic point of the book is just, there’s all this data out there; you should know about it.

Seth Stephens-Davidowitz: (17:11) It’s a self-help book with 50 tables and charts, which is not a normal self-help book. The other reason I wrote Don’t Trust Your Gut is I read more self-help than anybody I’ve ever known, which is a little embarrassing to admit as a supposed intellectual, because intellectuals hate admitting they read self-help, but I read everything like how to get more powerful, how to get richer; when I was single, how to date, fighting depression, because I struggle with depression a lot. I just read all the books and I’m always so frustrated because I feel like the books are just not really credible, in my opinion, I’m like, the evidence isn’t to the rigorous standard that I would think evidence should be put to. And so I wanted to write a book on all these topics where I’m like, I’m just going to give actual evidence on it.

Seth Stephens-Davidowitz: (18:05) And the other thing I did in this book, which was very important to me, is I notice a lot of self-help books — because I read so many self-help books — they have a theory (if they call themselves an evidence-based self-help book) they have a theory and then they just Google a study to defend that theory. There’s a study for anything, like “Broccoli is good for you. Broccoli is bad for you. Meditation’s good for you. Meditation’s bad for you. Three glasses of wine are good for you. Two glasses are good for you.” So if you have a theory, you can go and find a study to defend that theory and then say it’s “evidence-based.” What I did in this book instead was I literally had zero idea what I was going to say on any topic. And I just read every study in the area and I’m like, these actually I believe are convincing.

Seth Stephens-Davidowitz: (18:54) Usually I found there’s one methodology or one series of studies that is just way better than all the others, because it’s using like an enormous data set whereas other ones are using 30 data points. So I’ll give you an example: happiness. So, obviously, important question. Most happiness studies, even happiness studies that went viral are totally crap in my opinion. They recruit a couple hundred undergraduate students and they ask you what makes you happy or something. And it’s like, that’s a study. And the sample size is tiny, who knows how undergrads relate different than other people?

Satyen Sangani: (19:31) You need to look at their Pornhub searches.

Seth Stephens-Davidowitz: (19:33) Yeah. It’s just like, it’s so uncredible. And then I’ve found out about these projects. They’re just available thanks to our digital age, where they ping people on their smartphones and they say, “What are you doing? Who are you with and how happy you are?” And the largest of this “mappiness,” they have more than 3 million data points on happiness. And they’ve done all these amazing things that are just so much more convincing than every other study, they’re taking advantage of natural experiments. I always found there was one big thing that just stands out in the data. And the major lesson I found from these studies is that the things that make people happy are so freaking simple. They’re like things that human beings have been doing since we were hunter/gatherers.

Seth Stephens-Davidowitz: (20:13) People are happy when they’re with their friends, when they’re going on a hike in nature, near water, in a beautiful environment. And many of the modern aspects of life when people are doing them, they report they’re really miserable. So social media offers the least happiness of every leisure activity. And there have been experiments that show that people’s depression lowers when they are randomly chosen to quit Facebook. Things that make people happy are really, really simple and old fashioned. So, I conclude Don’t Trust Your Gut with what’s the data-driven answer to life. And I say the data-driven answer to life only uncovered in smartphones due to our modern era, big data, the data-driven answer to life, be with your love on an 80 degree and sunny day, overlooking a beautiful body of water, having sex —.

Satyen Sangani: (21:05) Yeah. And get off your, again, get off your damn smartphone.

Seth Stephens-Davidowitz: (21:07) Yeah. That is literally, that to me boils down everything in happiness research, the simplicity of the things that make people happy and the old fashioned nature of many of them.

Satyen Sangani: (21:22) So you obviously looked into happiness and you mentioned a whole bunch of other fun areas like parenting, give us a grand tour of … or not a grand tour. Give us the other top two things that stood out that were maybe counter to people’s gut and what they otherwise believe or might believe.

Seth Stephens-Davidowitz: (21:38) So most of how kids turned out is genetics. There’ve been all these studies of twins and adoptees and they found that parents on average make less of a difference than we usually suspect. But then recently they found in tax data that the neighborhood map kids grow up in matters a lot. They actually did something really clever. They compared families that move when they had multiple kids at different ages. So one of the kids is going to spend 10 of their years in Denver and another kid’s not going to spend those years in Denver. So what happens when they move to this area? Did the kid who spend more time in Denver do better off? And then you’re controlling for the genetics aspect of things because it’s the same family. And they found that neighborhoods tend to have bigger effects than people might suspect to the point that I think it’s pretty clear that the neighborhood you raise your kid in is by far the most important decision you make and swamps all the other ones.

Seth Stephens-Davidowitz: (22:36) And the reason that neighborhoods turn out to be so important is kids seem to be influenced by the other adults you expose them to, adult role models are really important. They’ve done studies that little girls who grow up around female scientists are much more likely to become female scientists themselves, and Black boys who grow up in a neighborhood with a lot of Black fathers around, even if they grow up without their father around have way better life outcomes, they have great role models around them.

Seth Stephens-Davidowitz: (23:07) One thing I take from this is that parenting is hard in part because kids rebel against some of the things you tell them. So the parent/child relationship, as Freud told us, is a very complicated one. But the neighbor/child relationship is much less complicated. So if the person down the street or your friends, your kids are probably going to think they’re really cool. So they may be having more influence than you sometimes think. So I think a data-driven answer to parenting is almost outsourcing parenting a little more than you might naturally expect, but putting your kids around in the environment of people you want them to turn out like.


Don’t Trust Your Gut

Satyen Sangani: (23:46) Getting back to the title of your book, with regards to gut instinct. So you have all these findings in all of these different areas of life where people would perhaps behave differently than the science might otherwise tell them. Do you find that lots of people ignore this data? Did you find that most of the simple advice is stuff that just people literally didn’t do or wouldn’t do or couldn’t do, and that their gut instinct was counter to many of these findings?

Seth Stephens-Davidowitz: (24:13) I think their counter. And I think telling people, it actually has an effect. There actually have been studies where if you tell people what activities tend to make people happy and there’s a randomized controlled trial, they do more of these activities and they are happier. So just telling people that watching Netflix and playing computer games, when you study 3 billion iPhone pings, tends to make people really unhappy, but taking a hike with friends tends to make people really happy. The evidence suggests that that could lead to differences in decision making. And certainly in my life, I’m a very different person than I was before I wrote Don’t Trust Your Gut, because I learned all these things that I hadn’t known that have led to different decisions.

Seth Stephens-Davidowitz: (25:00) There are some things you tell people, and it’s like, it’s not surprising them, and the problem is just doing it. So you can say eating processed foods is bad for you and going to lead to gaining weight, and don’t eat Doritos. And then I don’t think just telling people that Doritos are bad for you is going to lead to a huge difference in behavior. The problem is not with the knowledge, the problem is our brains are designed to crave these flavors that Doritos and other processed foods have in abundance. But I do think there are things like “relax about everything, but put more energy into the adults you expose your kids to,” or “be an extreme version of yourself and ask more people out,” or “take more hikes and try gardening, or spend more time in nature and spend less time on social media and computer games” where I think people — I hope and believe that some people just being told that will make different and better decisions.

Satyen Sangani: (26:06) Yeah. That is not true of one of my kids, but certainly we are going to keep on trying and I’m going to put your book in front of him and see if that makes an impact, or maybe I should have a neighbor put a book in front of him and see if that has an impact.

Seth Stephens-Davidowitz: (26:26) There are some things in the book like, I have a section [that] basically [says] looks are massively overrated in dating. And I literally titled that section: “Looks Are Overrated and Other Advice You’ve Long Been Told and Consistently Ignored, but May Be Slightly Less Likely to Ignore It.” If you look at long-term couples, they’ve done studies of more than 11,000 couples, and what predicts happiness, and the conventional attractiveness to your mate has basically no predictive power. The things that predict happiness: Romantic happiness is pretty hard to predict, but the things that have some predictive power — psychological traits, like someone having a growth mindset or being conscientious, having a secure attachment style — all these psychology terms that I’ve ignored in my life have a lot of predictive power and things like a conventional attractiveness, or how tall they are, what occupation they’re in, all the things that basically everyone tries to date on, have no predictive power of how happy you are.

Seth Stephens-Davidowitz: (27:25) So clearly if you want to be happy long-term, not only should you care less about looks, you should actually focus more of your energy on less conventionally attractive people, because the competition is so much lower, because everybody so overvalues looks in dating that even if you do end up with somebody who’s beautiful, you may have to sacrifice a lot because everyone’s trying to date someone beautiful. Maybe they have really bad psychological traits, which is the reason they remain single or something. But again, I’m telling people that, and I have no idea if anyone’s going to follow that, because that’s probably also hardwired in our DNA to seek out conventional attractiveness.

Satyen Sangani: (28:06) Well, there’s that great scene from A Beautiful Mind where John Nash is at the bar. You’ve probably seen the movie because you’re an economist.

Seth Stephens-Davidowitz: (28:14) Yeah, yeah. The game theory on how to win a beautiful woman or something.

Satyen Sangani: (28:17) Right. Right. And the idea was go for the — all of us should go for the least, or the not most beautiful women because in theory then there would be a chance for all of them to win. But that’s hard to do obviously.

Seth Stephens-Davidowitz: (28:33) Yeah.

Satyen Sangani: (28:34) And it would require, well, it would require knowledge and it would require discipline. So I guess, a lot of your impact, or at least the impact that you’re trying to make, and I think it’s very similar to this in some ways, the impact that we’re trying to make, which is how do we get people to think more scientifically? And I guess, as you’ve done this work, both with, obviously Don’t Trust Your Gut has not yet come out, but how have you seen impact come back to you? And do you think that this is now, this idea of scientific thinking is becoming more popular or less popular? Are we all fighting an uphill battle? Any learnings there before we move out?

Seth Stephens-Davidowitz: (29:10) It’s become more popular, but the difficulty I’ve seen is that people want to do what they want to do. And then they think that being scientific is finding evidence to confirm what they already want to do, which is not, in my opinion, the scientific process. The scientific process is like how I approached Don’t Trust Your Gut. You could read my book proposal. I’m like, I’m going to write about parenting, I have no idea what I’m going to say. I’m going to write about entrepreneurship, I have no idea what I’m saying. Write about happiness, I have no idea what I’m going to say, and then allowing yourself to be surprised or, and then make different decisions based on that. And that’s very hard. What I encourage in a data-driven mindset or an evidence-based mindset is to be willing to go where the data tells you, even if it’s really, really shocking or really, really surprising.

Seth Stephens-Davidowitz: (30:00) There was actually … when I was doing my research writing my New York Times column, I did this study of Stormfront, this white nationalist site, getting back to depressing topics. And I talked about how many American anti-Semite there were on this site. I just found out, my mom told me that my dad, when I wrote this column, his response was, “Seth has gone crazy.” And he’s like, “Seth has gone down a wrong path of conspiracy theories, thinking that there’s some group of anti-Semites in the United States of America these days, Seth’s lost his mind.” And then the Charlottesville protest happened where all these people were chanting “death to Jews.” And my dad went to my mom and he was like, “I guess Seth was right.” And I think the real data-driven mindset is when you have data that makes you look crazy, to be willing to go with the data, I guess, against your gut. So I also thought it was nuts that there were these secret anti-Semites in the United States. I certainly never encountered them despite being a Jew in the United States.

Satyen Sangani: (31:06) Yeah. Which is, I think, a phenomenal learning. If you’re a leader trying to build an organization that is trying to bias towards data, then look for people who have taken brave or non-conventional tacts, or at least at the very minimum, don’t tie their ego to the outcome with whatever judgments that they happen to make. I think that’s a really powerful thing, because often if you have somebody who’s really outspoken and forward, they’re going to be perhaps tied to the thing that they were arguing for or the thing that happened to have been popular or the thing that got them some notability.

Seth Stephens-Davidowitz: (31:47) The other thing you don’t want is just yes people, who you have a theory and you tell your data scientist, “Is my theory right?” And the data scientist just comes back and says, “Yes, yes, yes. You’re right,” no matter what you said. You’ve got to be willing for the data scientist to come back and say, “No, actually your theory is wrong. We need to go a very different direction.”


Building a Career Around Science

Satyen Sangani: (32:07) Yeah. Which I think is true for leadership at almost every level, but certainly it’s got to be true of the people who are purportedly responsible for science within your organization. So you’ve obviously done the scientific work. What advice would you give to those folks who are starting out and want to be able to build a career in and around science? Any learnings that looking back you would’ve done differently or the same?

Seth Stephens-Davidowitz: (32:29) I think being early is very valuable. This does go with Don’t Trust Your Gut. One area where our guts are terrible is exponential curves. So when something’s growing exponentially, we can’t really project out, and getting in early on an exponential curve is really, really valuable in your career. And it goes to, a little bit, the same thing as the dating advice where you have to be willing to look a little weird. So when I was first studying Google searches, when I’m like, I’m going to devote my career to studying Google searches. I didn’t devote my whole career with it; but I devoted 10 years. people are like “Seth’s a freaking weirdo.” And I told you, my dad thought I was crazy.

Seth Stephens-Davidowitz: (33:09) Now it’s considered pretty obvious. Again, I go to a talk and say, who’s heard of Google Trends? 95 percent of people raise their hands. They have a Google Trends analyst on the team, but when I was doing that, it was considered insane, and people are like, “This isn’t properly weighted data. And there are all these flaws in the data,” and they’re like mentioning all the flaws and I’m like, yeah, it’s not perfect data, but the fact that millions of people are typing N-word jokes into Google is clearly revolutionary data that we have to be able to get some insight out of.

Seth Stephens-Davidowitz: (33:42) So I had this very strong feeling that it was going to be big even though at the time it wasn’t big. So I think that’s something to try to do in your career, which is hard for people. Maybe I should write a book, Don’t Follow the Herd. If everybody goes left, people’s instinct is to go left. If everybody goes left, my instinct is to go right. And I think that can be… it kind of goes to not dating the person everybody else wants. Not trying to date the person everyone else wants to date. I think that’s a valuable life approach.

Satyen Sangani: (34:15) Yeah. All the strategies about how do you focus your scarce time and resources and energy? And if you’re doing what the average person is doing, you’re just going to get the average or maybe even the below average outcome. So, that’s incredible advice. Seth, this has been just awesome and such a fun episode and such cool writing. So thank you so much for taking the time and for giving us everything you’ve learned over the years.

Seth Stephens-Davidowitz: (34:43) Yeah. Thanks. I only have one joke left in life and it’s that I wrote the book, Everybody Lies, and there’s nothing people lie more about than how much they enjoyed your time on the podcast, or how much they enjoyed your book, or how much they enjoyed your talk. So I am highly skeptical of people’s compliments on these topics, but thank you.


How We Collect Data Impacts Our Findings

Satyen Sangani: (35:03) Yeah. I lie about a lot of other things apparently, but not in this case. That’s so awesome. Thank you so much.

We use data to bring objectivity into a world that isn’t always objective. How we collect data impacts our findings. Human bias can sneak in when we choose what to collect and how. Worse still, we can’t always trust the data that we do collect, and it doesn’t stop there. How you present data will impact your response and further change what you’re measuring. Nuanced qualified claims rarely peak interest. You’re better off picking one detail and diving in to make one big sensational claim, which then introduces more bias, more reactionary behavior and more distortion. Thanks to Seth for unpacking the lies behind the confusion behind what we think might be the reality. And thanks again for a super amusing conversation. This is Satyen Sangani, CEO and co-founder of Alation. Thank you for listening.

Producer: (36:08) This podcast is brought to you by Alation. Are you a leader in data? There is a new white paper detailing the steps you can take to launch an effective data governance initiative. Visit alation.com to get your free copy.