Transcript
Greg Stilestra:
Hello and welcome to the Behavior Change Podcast by Lirio, the program where we explore the marvels of behavioral science and ways of applying it to make a better world. I’m your host, Greg Stielstra. On today’s show we’ll talk with behavioral scientist Evelyn Gosnell. She’s an author, speaker, consultant and managing director of Irrational Labs in San Francisco, where among her other duties she leads their BE Bootcamp. She has a BA in cultural anthropology from Duke where she graduated summa cum laude and a masters in marketing. She speaks fluent French and Russian, but we’ll stick with English for our conversation today. In a moment we’ll talk with Evelyn about a wide range of behavioral science topics including why we struggle to keep New Year’s resolutions. The power of fresh starts, something called Ulysses contracts, how the wrong metrics can mislead, and the benefits of the BE bootcamp that she leads. But as always, we’ll begin with the bias brief.
Bias brief number 178: Commitment and Consistency
People often say that past behavior is the greatest predictor of future behavior, but why? It’s because our actions, beliefs and commitments help form our self-perception, and how we see ourselves guides our choices. If you want to change someone’s behavior, one method is to change the way they see themselves and you can do that simply by getting them to make a commitment. That’s what Stanford researchers, Jonathan Freedman and Scott Frazier discovered with an experiment they ran in 1966. In a neighborhood in Palo Alto, California, researchers claiming to be from the community committee for traffic safety were sent door to door asking homeowners to sign a petition promoting safe driving. Some homes were skipped to create a control group. Two weeks later, a different person visited each house claiming to represent a new group called Citizens for Safe Driving. They asked homeowners if they would install a very large and unattractive sign in their front yard that said, drive safely.
The results were revealing. With the control group, the homeowners who had not received the first visit or signed a safe driving petition, only 17% agreed to install the drive safely sign. Of the homeowners who had received a visit and who had signed the safe driving petition two weeks prior, 76% reinforced their commitment to safe driving by agreeing to the sign. As it turned out, the homeowners view of the drive safely sign depended on their view of themselves. While most of the homeowners probably looked at the sign and thought, “Do I really want that thing in my yard?” An easy no for those who had not signed the petition, those who had signed the petition were motivated by a different question. What would a person who is committed to safe driving do? That person, the one they had become by signing the petition, would accept the sign and so they did.
You can apply commitment and consistency using something called the foot-in-the-door technique. Start by persuading the target to agree to a small request. When they do, they will be more likely to comply with a larger, related request. Like in the example when someone signs a safe driving petition, they are most likely to display the oversize drive safely sign in their yard. Here are some factors to enhance the effectiveness of the foot in the door technique. Align your requests. For the technique to work, the initial small request must be related to the large request. Signing a safe driving petition and putting a safe driving sign in your yard work because they’re related. Signing an anti-litter petition and putting a safe driving sign in your yard are unlikely to work because they are unrelated.
Require action. The small request is more influential when it requires an action. Asking people to sign a petition is more influential than say, asking them to state their opinion. The more effort that goes into a commitment, the greater its ability to influence future behavior. Make it public. Commitments are more powerful when they are public. You could amplify the influence of someone’s signature on a petition, for example, by publishing the petition online for all to see and phrase it as an inner choice. Whenever possible, phrase commitments as an internal choice rather than an external restriction. “I don’t litter.” An internal commitment is more powerful than. “I can’t litter.” An external restriction. Avoid rewards and threats too because they can add external pressure, which may reduce commitment. Whether you’re looking to increase patient adherence, boost energy efficiency program participation or help a loved one achieve a desired goal. Commitment and consistency bias can be a helpful tool for achieving better outcomes.
GS:
And now my conversation with Evelyn Gosnell.
I wanted to start by saying happy new year Evelyn and thank you for joining the podcast.
Evelyn Gosnell:
Happy new year to you and yeah, thrilled to be here and to chat nerdy behavioral science things with you. It’s my favorite thing.
GS:
It’s interesting. I worked with a gentleman named Dan Buettner, who studied the healthiest places where people live the longest on the planet and some of the tricks he talked about for curbing inappropriate eating were putting snacks in a separate drawer labeled snacks. So you don’t see it when you open up the regular food pantry. You only see it if you open up the snack drawer and you’re reminded that it’s not the healthiest food because of the label.
ES:
Yeah. On my phone if you scroll, on the main screen, it’s not available, but you have to kind of scroll several screens past and then there’s a folder that’s called Waste Time and then Instagram and Facebook are in that folder. So I’ve created friction that it’s not on the main page so I have to scroll a little bit. And then there’s this sort of emotional trigger of like, I feel like I’m going to waste time. But again, the ideal solution really would be to remove them entirely from my phone. One time, actually I did do the password thing, I had a paper that I was working on, it was a Christmas time. Everybody else was having fun and I had to sit at home and work on a paper and it was just perfect scenario of just to like, “Oh, let me just check a little bit and spend a little bit of time and see what everyone else is doing.” And it was just not effective. So I had to give her my password. She changed it. I could physically not access Facebook until I finished that paper.
GS:
And you finished the paper.
ES:
Yeah. There we go.
GS:
You knew how to manage yourself. Now we’ve been giving New Year’s resolutions a bad rap, but there is some evidence that planning to start something new, make a new commitment around a date like New Year’s can actually help us succeed temporal landmarks, they’re called.
ES:
Exactly. And I think that it is just a moment I went on a hike on New Year’s day with a friend and was talking about this and this exact topic of yes, New Year’s resolutions. It’s kind of an artificial time frame, but it is a trigger point for us to kind of pause and take stock of our lives and say yes, if there were a moment to change, why not now? So I think it’s really nice… Yeah, the fresh idea, fresh start effect is a beautiful concept that I think, again, even companies should use more. New Year’s is sort of an artificial one who’s to say that January 1st really it doesn’t actually matter, but it matters in our minds and that’s what’s significant here. And so when people when it’s their birthday or, and that’s especially true when it’s a milestone birthday or all kinds of things like first day spring.
GS:
Yeah. First day of spring, first of the month, a Monday, you’re likely to have more success just picking Monday as the data begins to pick during the week. Well, I’ve read that part of the reason for this has to do with how we view our present self differently from our future self. Our future self we think is pretty infallible, pretty amazing. And that if we start something on January one, that individual, that version of myself is much more likely to be successful because once I crossed that temporal landmark and sort of freed from all the foibles and mistakes of my past and present.
ES:
Absolutely. I think one of the details there is when we’re thinking about our future self, the more we project, the farther away the future is in that case. Right? So if we’re thinking about our futures health being tomorrow, that’s one thing. But if we’re thinking about our future self being a month from now or a year from now, the longer the temporal landscape is, the harder it is for us to think through all of the details that might get in the way of doing X, Y, Z thing. So right now in our present moment, we can think about, “Oh, I can’t go to the gym tonight because I have these meetings and they’re going to run late, or maybe there’s going to be traffic or I’m a little bit tired from work.”
So all of this stuff is salient and relevant to us. And so it’s relatively easy. We can think about what those are, but in the future we’re less good at thinking about all the things that will likely get in the way. So it kind of goes back to optimism and just not thinking through all of the barriers that are going to happen.
GS:
And because we are not as able to anticipate all of those barriers, we become overly optimistic about our chances for success.
ES:
Again, I do think that I don’t want to write off New Year’s resolutions entirely, but I think it’s just how we design the change and designing in a way that increases the chance of success.
GS:
And that’s your area of expertise. Our failure to make good on our intentions is not limited to New Year’s resolutions. It’s this intention, action gap is something people struggle with all year round. Talk a little bit more about that and some of the ways we can help close it.
ES:
Every domain, if we think about the areas that are most important, so Irrational Labs works in the domains of health, wealth and happiness. So what products and services are out there that it can improve people’s health, wealth and happiness. Those are the projects that we take on and the domains that we’re most interested in. And really those are they’re the hardest, right? So if we did a behavioral, we probably wouldn’t even need to do a behavioral economics bootcamp for somebody who works at Coca Cola and is trying to sell Coke or somebody who is selling any kind of thing that has a hedonic benefit that is in the present moment. That job is much easier. Most of us, however, are working in domains where the benefit is in the future and the benefit is functional, right? So retirement savings is not a hedonic benefit, at least not directly. In the long run if you do it correctly, there can be hedonic pleasures that are tied to it. I guess you could argue.
But overall it’s a very functional thing to say, how am I going to ensure that I have enough money to have housing, to have food to eat, to be able to support myself. And so that is what we do. We focus a lot on, and those are hard nuts to crack. And so what that means often is that it’s not… I wish that we could say yes, you just look at behavioral science and the solution is the obvious solution is right there in front of you and just do that. And if that were the case, everyone who worked in those domains would have already done them. The reality is that human behavior is complex and getting behavior change in these domains is a hard thing.
And what that means is we often have to try a lot of things and experiment and fail a little bit along the way and learn. And a lot of times it’s not just one intervention. Maybe one intervention works over a period of time and then it loses its… That’s true in many health interventions. It’s hard to find one amazing solution that works forever. It’s just unfortunately humans are, I guess fortunately and unfortunately humans are more complex than that.
GS:
So why do you think that is? Why do you think that particular interventions lose their effectiveness over time?
ES:
Again, I think that’s hard to answer generally. I’m thinking of off the top of my mind on a lot of the interventions around, for example, exercising more. So when you add certain incentives you can get people to exercise more. Katy Milkman did a nice one where she called it temptation bundling where people were only… They got audio tapes. You’re probably familiar with this study. They got audio tapes of the hunger games, so she found something that is highly hedonic, right? Highly juicy.
GS:
For some.
ES:
Well at the time when it first released, now you would want to pick something I don’t know what the modern version, the more updated version of that is, but she picked something like that and had people either access the audio tape at the gym while they’re exercising versus just having free access to it. And what she found was that it was more effective to pair it concept of temptation bundling is taking something that isn’t immediately pleasurable, but it’s useful in the future. Right. Going to the gym and pairing it with some benefit that you care a lot about. So you’re pairing something kind of yucky with a nice reward. But I think, again, if you think about that example, how long is that going to last? What’s the next thing that you need to do after that? So that’s where it’s just challenging.
GS:
So what you’re describing is sort of the importance of being self-aware and in particular aware of the ways that we operate psychologically so that we’re prepared to help ourselves be our best self. Most people aren’t, I mean, they’re not aware of some of the science that you know. What’s a good way for a lay person who wants to understand themselves better to get the one on one?
ES:
Oh, that’s an interesting question. I didn’t think you were going there. I was more thinking about, for me, the lack of general understanding of these behavioral science elements puts all the more responsibility on us for those of us who are working in these products to design them well for people. To design them in a way that helps them close their intention, action gap, to align with their best with their longterm interests. So I think that’s-
GS:
Let’s go there because I think that’s also a good question, which is product designers, marketers need to be aware of these things in order to do their jobs effectively. And yet many aren’t. Richard Dailer said there’s no such thing as a neutral presentation. So everything these people are doing has an effect, it seems to me they’re obligated to understand what that is so that they can do no harm and they’ll be better at their jobs.
ES:
Yeah, I think it is. So the question that we’re talking about is understanding user’s intent or user’s best interest. And I think in some cases that’s more obvious than others. So let’s take meditation for example, which is a wonderful thing. It’s also very hard to do if you’ve not been a meditator and you know the research of the positive elements of meditation and you want to add those to your life. But all of a sudden going from zero to not being a meditator to being a meditator is a hard thing. So let’s say you download a meditation app, you pay for a subscription, there are many wonderful apps out there. So you do it the first day because it’s all exciting. Your motivation is super high, maybe the second day and then from the third day on something else happens because it’s not a habit you have no built in routine around this.
So I think that that’s the open question is what is the responsibility and where is that line for the app developers to say we are going to help you design a world where you are more likely to do this? So for example, in the signup flow, should we get them to think about, have a kind of implementation intention prompt around let’s think about… Because their motivation when they’re first signing up is actually likely very high which you have to consider, should we have them think about, let’s think about the day, when do you think is best for you? Or actually from a lot of the meditation research it is actually better to do kind of first thing in the morning. So should we prompt them to say usually you wake up at 7:00 AM, but how about we set your alarm at 6:50 instead?
So we’d build in the 10 minutes. So you’re kind of thinking through the probability and the likelihood of the next day and the next day after that. And how do we build that in to the success versus, I think what happens too often is as product people, we live in the world. Imagine you’re a product manager in meditation, you live and breathe meditation, you know the importance of it. Of course, meditation’s important. So we take that and we extend it to our users way too much like, “Oh yeah, they’re downloading the app, they want to meditate. Clearly they’re going to follow through and they’re going to meditate every day.” And they don’t. So I think that’s where it’s our responsibility but there is a line, if they don’t meditate are we pinging them with notifications, to what extent? That’s where it gets a little bit tricky.
And again, I think with something like meditation it’s a little bit clearer that that was the user’s intent. I think with other things it gets a lot messier. Imagine various social media apps or things like that. Games that suck people’s attention. We should not just take people spending time in the app as a proxy for that was their intended goal and they enjoyed it. There could be many instances where people are spending time in something but they later regret, oops, I didn’t mean to spend, I thought it was just checking this and opening it and I got sucked in and that was not their intent. So we have our responsibility as product folks to think deeply about and in some cases it’s appropriate probably even to ask like to align on the user intent.
GS:
Do you think users know what they want in many cases?
ES:
I think that’s super context dependent, but I mean we could live in a world. You could ask me, I mean Apple is kind of going in that direction with screen time limits to say like, well, I want to tell you right now in a cold, rational state, I want to tell you that I only want to spend 10 minutes on Instagram a day. So there you are capturing the user’s intent. So imagine a world right now, screen time limits are kind of weak in that you can just get around them. Imagine a world where I wanted to really create a true Ulysses contract where I was blocked. It literally said, I’m committing right now. I only want 10 minutes on Instagram a day. Do not give me the option to go around this. Some people might opt into that.
GS:
Tie me to the mast.
ES:
Yeah, it’s a worthwhile question and maybe they’re testing that. I don’t know. That would be fascinating. I think those things should be tested.
GS:
It does seem to require product managers to take a long view rather than a short one. Right. So in the near term I may want people engaging with my app as much as possible, but if in the long run that causes them to become disillusioned and leave, I’ve lost.
ES:
Absolutely. I think that’s a wonderful point. And I think something that we underestimate and I think that the answer to that is metrics. I think that teams we are what we measure and if you’re just looking, if you’re a team every day and your daily stand up, if you’re looking at what was engagement today, how many people opened the app and that was the only metric that you’re focused on, that’s absolutely the wrong focus. And I think that we need to be getting teams to think more about long-term metrics and having those be equally, if not more important than the short-term metrics.
GS:
So you raise a very good point. I’ve noticed that quite often people will report anything they can record, which creates a clutter of metrics, some of which are important, others of which aren’t. What’s a good way for someone to discern which metrics they should be paying attention to, which they shouldn’t? And I know it’s hard to answer that generally, but maybe I’m asking how do you think about the user and identify metrics that align with their goals?
ES:
Yeah, I think it is so context specific, but I mean I think you just start at the highest level possible and back down into it. I think that measuring in the bootcamp we talked a lot about key behaviors, you can’t drive behavior change unless you’re measuring behaviors. So a thing that we see way too commonly in many companies is people are like, “Yeah, my job is to increase engagement.” And okay what exactly does that mean? And we’ve had instances where we’ve gone into companies and we split up the team and the leadership team and all of them say separate things.
So I think really thinking about the behavior and designing for that. So back to the meditation app, opening the app is not very relevant. Right? That increases your engagement score because they opened it, but we might want to have a more clearer goal of meditating at the same time in the morning for five days in a row. And as soon as you define a metric like that, the team then aligns much the actions that they’ll take, the product changes they’ll drive are very different than if you just said, just if opening an app counted.
GS:
Okay. Where you end up helping the user achieve his or her goals.
ES:
Absolutely.
GS:
Zig Ziglar said, you can have anything in life you want if you help enough other people get what they want. Maybe it’s time we put ourselves in our user’s shoes to a greater extent.
ES:
Yeah, absolutely. Because I mean then I get that there’s profit motives in all of that, but that makes sense. Right? If we get someone to become a regular meditator and the benefits that they incur from that, they will resubscribe. They may have other friends of theirs that they might promote it to them. So I think there’s very much a world where there’s win-wins here.
GS:
Let’s shift gears and talk about the world that you live in, which is the bootcamp. What is that and how can it help product people apply these principles to their product and make them better?
ES:
Yeah. Great. Thanks for letting me kind of nerd out and talk about it because it’s one of my favorite things. It gives me great joy to be able to share the insights around how humans behave and how we make decisions and some of the irrational ways we approach things and really understanding that kind of body of work and allowing product folks to say, “Okay that’s wonderful, but how does it apply to me?” So I think that that’s one of the things that Irrational Labs does really well is saying there’s all of this research out there and no one’s stopping you. Right? If you wanted tomorrow you could just go read all of the academic papers out there and lots and lots of books. But I think what a wonderful opportunity is to say great, I understand these concepts and they’re presented to me in a way with frameworks and case studies and specific examples that help people understand, “Oh okay so I get these concepts and now I have a very clear action plan, a three step process of what I can do differently.”
So it’s very much of an applied session on or multiple sessions on applying behavioral science. And maybe, I mean I’m sure listeners would want to hear your thoughts and experience as well.
GS:
Yeah, so what we’re talking about is do you consider it nine or 10 weeks, Evelyn?
ES:
We now have three versions. So we have a 10 week version, which is our full bootcamp, which is what you went through. And then we have two shorter versions. We have a just a two day sort of intensive intro and then we have a four week as well. So we have all of these versions for different interest levels and price points obviously.
GS:
And you learn to basics behavioral science, what it can and can’t do. You learn about many cognitive biases that describe people’s behavioral tendencies. And most importantly for me, I learned a framework for applying that knowledge to a particular behavioral problem that you’re trying to solve either with your product or in your business. That 3B process as you call it was super helpful for me because it gave me a way to go attack a problem, come up with a solution, test it through experimentation and arrive at a really good solution.
ES:
Yeah, I mean, we put a lot of time and energy into thinking about how to make it very actionable and just useful for product folks in terms of driving specific changes. So yeah, you mentioned the 3B framework. We also taught folks how to do a behavioral diagnosis, which is mapping out your entire product flow, identifying the behavioral biases that are at play and coming up with concrete solutions that you want to drive. And then we also had a whole segment on experimentation because I think many people kind of get excited. I think there’s kind of a split, there’s like excitement around experimentation but also around a lot of fear saying, “Oh that’s so complicated, or we’re not large enough to do that, or that just seems like a lot.”
And so they’re a little bit afraid. Some companies, many are doing lots of experiments. But I think there’s this sort of split and so that’s why we taught folks sort of the basics of experimentation, the do’s and don’ts, how to have the control, how to have the intervention, what is the sample size that you need, all of these important things on how you would actually want to run an experiment. And we had folks design them and present them and many of them ran experiments and presented final results, which was super exciting to see.
GS:
Yeah, I think the experimentation was a really important element because you think you know but you don’t always know and you’ve been exposed to so many different experiments and outcomes. I’m wondering whether there are any particular experiments you’ve run where the outcome was surprising or it didn’t match your hypothesis.
ES:
We actually do this a lot. Irrational Labs we’re a small team, but we take bets on what the results are going to be and we’d bet each other lunch. Richard, who you know who teaches our experiments segment is particularly into betting and so we have a very complex matrix of who owes whom lunch.
GS:
Did Richard buy your lunch in Paris?
ES:
He did not. I should have. That would have been a good idea. I’m trying to think of examples off the top of my head. There’s definitely been plenty of instances where we’re either wrong or we just got the level of impact kind of off. So sometimes it can be dramatically more effective. I think what happens a lot in experiments is you sometimes actually don’t see, if you look at your full data set, you don’t see any results, any significant result. But then you can do a split and you say, okay, let me split this between men and women and then all of a sudden, wow, there’s no effect on men, but there’s a kind of strong effect on women or women under X age. And so then your solution then becomes a little bit more complex.
But I think that that’s, it’s kind of the more advanced version of experimentation is to really the gender question might not be relevant in all contexts, certain products that it’s completely relevant. But in some it might matter. In other cases it might be a question of income or it might be a question of experiment experience with that product. Old users do this, experienced users do this and newer users do that.
GS:
I also find that experimentation is especially important in the behavioral science world because so many of our behaviors are irrational and therefore defy our intuition. So we sort of think it’s going to go one way. And then you discover it through experimentation that the behavior is different than you anticipated.
ES:
Yeah. And that’s why it’s so important to measure and measure the right thing. I thought about this going back to Paris, I thought about this because I was in a park and there was a little sign, a little cigarette, I guess little trashcan, but it was transparent and there were two you could basically vote. And it said which of these two bridges do you like in Paris more? And there were two different names of two major bridges in Paris. And I thought so their goal, the designer, the choice architect there who created that is thinking, oh, this is an intervention. I’m guessing here she designed it in a way to think to reduce litter, which is a nice effect or a nice concept. However, if you were just walking through the park and let’s say you’re trying to quit smoking or trying to reduce smoking, you weren’t planning to smoke, but now you see this and you want to vote, there could be a backfire effect there where-
GS:
You have to take up smoking in order to cast your vote. That seems like an egregious selection bias too, you’re only getting the answers from smokers.
ES:
So I thought about it a lot, because that’s what you do as a behavioral scientist. You’re just walking in a park like a normal person with your family, but then you’re kind of going down this rabbit hole. So then I started thinking, well, my first reaction was it’s kind of weak. Like who cares about these Paris bridges? Not a big deal. And thinking more about it though, I think it’s actually, in some ways it’s better because imagine if you did something with much more, imagine that it was like, who do you like better? Obama or Trump? My hypothesis would be that you would increase the rate of smoking and unintended or unplanned a cigarette in that case.
GS:
Just so you’d have the ability-
ES:
Just because you can vote, you care much more. And so maybe the bridge thing, it’s a weaker or less passionate things, but maybe that makes it a better intervention. If our goal is to reduce littering and not increase smoking, so-
GS:
Maybe they should cast their vote with granola bar wrappers instead-
ES:
I don’t know if they eat granola bars there.
GS:
Well, hey, some of the people listening to this right now might want to attend a bootcamp. Experience what I did, benefit from your good instruction, how can they do that? Where can they learn more?
ES:
So our website is bebootcamp.com and they can learn all about the boot camp program again. And we have three different offerings, a two day and we have several two days sessions that are already planned. We have a four week and we have a 10 week one. So bebootcamp.com.
GS:
I will add, not only is the education great, but did the experience with the other students is exceptional. You’ll be joined by a lot of smart people who are also interested in behavioral science and applying it to do something good in their particular corner of the world. So it’s very worthwhile experience. Evelyn, we’re out of time for today, but I’d love to have a conversation with you again in the future sometime if you’re willing.
ES:
Of course. Yeah. You know me. I love talking as long as about nerdy behavioral science stuff, I’m on board.
GS:
You’d always would be. So thanks for joining us. We’ll talk next time.
ES:
Wonderful. Thank you.
GS:
You’ve been listening to the Behavior Change Podcast by Lirio. Lirio provides an email based behavioral engagement solution that uses machine learning, persona based messaging and behavioral science to help organizations motivate the people they serve to achieve better outcomes on the web at lirio.co L-I-R-I-O.C-O or follow us on Twitter @lirio_llc.
© Lirio, LLC