Behavioral Science with Lilly Kofler

by Mitch Turck

Our latest episode of the Question Authority podcast features Lilly Kofler talking to us about behavioral science.

Listen below, or grab it wherever you get your pods. Want to hear more? Each new episode brings in the best & brightest marketers to teach DTC brands about the art & science of questions. Subscribe for more, and let us know who you'd like to hear on the show.


Mitch (00:05): Welcome to Question Authority, where the best and brightest marketers teach brands about the art and science of questions. Today, we're asking about behavioral science with Lilly Kofler.

Mitch (00:28): Hey.

Lilly (00:30): Hello.

Matt (00:31): How's it going?

Lilly (00:32): Good, how are you?

Matt (00:33): Awesome. Yeah, we're super excited to have you.

Lilly (00:36): It's a pleasure to be on the podcast, speaking to all of the wonderful brands that are part of the Shopify ecosystem.

Mitch (00:43): Well, Lilly, happy to have you on once again. So today you are talking about behavioral psychology and how that's applied to marketing and communications programs. Is that accurate?

Lilly (00:57): That's absolutely right. And what I like to say is that what I'm really good at is blending the science of human behavior with the art of marketing and communications to tackle challenges big and small.

Mitch (01:09): How would you define like where you work usually or where people like you tend to work in a corporation or in an agency or any kind of space? Where do you guys fit in, or do you not fit in and is that part of the appeal?

Lilly (01:22): I think there's an argument to be made for a little bit of both. In my professional experience, which sort of started in academia doing laboratory research specifically on kind of how people make decisions about value over time, to then working sort of in the agency capacity in broadcast advertising and digital, and now I have the pleasure of working in-house. I sort of work in decentralized finance and blockchain and a product that's on our blockchain is an app called Valora, which essentially turns crypto into usable money in all these delightful ways.

Lilly (01:58): So when you're talking about kind of the agency landscape of behavioral science, it very much exists as a capability, right? So it looks like a different way in, a different point of view, an emerging approach into tackling the brief that are oriented on a uniquely behavioral challenge. And in an in-house capacity, it's definitely a choose your own adventure or for brands and businesses, a lot of folks with this training sit in innovation hubs. Oftentimes, it comes to life as a UX designer or as a strategist who just has a different type of training. And as the world changes and as marketing and comms starts to look ever more digital, there are just more opportunities for a diverse background and skillsets, and one of which that happens to be very effective is training and human behavior.

Mitch (02:49): Yeah. That's what I find so interesting is that it's kind of like, as you pointed out, it's still very nascent to be using behavioral psych and human behavior in all manner of building your product and your brand. And so that's always kind of been my experience is that if it's not inside an agency who gets paid to do that kind of thing and on a real thorough soup to nuts basis, then it's usually sitting in the path of least resistance at a brand or something, right?

Mitch (03:18): Oh, where are your people who actually study your customers? Wherever the leash is longest, basically. It could be someone in product. It could be people in marketing. It could even be customer service, if they have a really strong manager and they have good performance metrics, then they get a lot more leverage to do what they want. But it is always funny to see where like from brand to brand, as you skip along like, oh, where are the people like me?

Lilly (03:43): Totally. And the kind of awakening of behavioral science really started in the 1970s, right? It's really young compared to what these folks are doing in the real world. But what I always say is that marketing and communications and design professionals, they're sort of the original behavioral scientists. They tested this stuff. They went into market, they learned, for sale signs should be read. They learned that without the neuroscientists, then kind of going improving it. And in my sort of experience in the applied world, what I've found is that behavioral science just gives a language and a framework for how to think about this world. Somebody came up with an insight from customer research of, oh, people are afraid to switch their brand because they don't want to regret that decision. When behavioral science land, I would call that anticipated regret. Same concept, same insight, sort of pushing a data-driven mindset, putting the customer at the forefront of everything.

Lilly (04:45): And often times, my role at agencies and even somewhat in what I do today is just being the library and of all that great research, right? When data, it feels like such an important commodity and it is, and especially smaller brands, younger brands, oh, we don't have enough. We can't afford it. There is so much data about your customer out there that you can get in nimble and agile ways. Inquiry is one of them, but also this incredible literature of desk research is also there waiting for you, which unfortunately you might then have to hire an academic because it is very much written by academics before academics, but the insights most certainly apply to people like you and me.

Matt (05:27): Here's a question too. You mentioned desk research. How would you go about your, let's say you're a brand, just kind of given our customer base, where would you start, if you wanted to start doing, as you refer to as desk research?

Lilly (05:40): To think like a behavioral scientist, you want to start with a problem, right, because getting that problem really, really targeted, really specific is then going to enable sort of these tactical approaches like doing desk research. Google is a fantastic tool. You can plug in questions big and small. If you then want to drill down a bit, that's when I would go from open door, regular Google to Google scholar. And what you want to do, this is my hack, is you search things like decision sciences psychology, choose your key word of the day, and then you search PDF. And what Google scholar will do is it'll sort of scrape all of these papers and the ones that are sort of free PDFs are available online. There's on the right hand side of the screen, it'll say PDF and you click that and it sort of circumvent some of the firewalls that get people out of this content.

Lilly (06:34): And then just start reading, skim the abstract. If it feels good, do your best to work through the intro. I would skip all of the methods if I didn't have an academic background and just go to the conclusion. Look for papers that are more like literature reviews or qualitative because a lot of stuff is very experimental and it's about the experiment, not the outcome, and just have fun, follow the citations, right? Maybe that paper isn't perfect, but they said a line in the abstract that really applied to you. They cited another paper, go Google that paper, find it on Scholar. There's probably a free PDF and find the joy in the search.

Mitch (07:12): That's a good way to phrase it there, Lilly, because I think a lot of folks would be surprised that, I mean, if you're willing to go down a wormhole in Wikipedia or somewhere else on a topic, then yeah. As much as this stuff might seem dry to a lot of folks who aren't used to it, to your point, I guarantee you'll start reading things and think like, wow, I never understood. I never knew that that's potentially is the way people behave, and then to start seeing the citations and the references to say like, okay, this is getting really interesting now.

Matt (07:44): I think as a founder of a lot of these direct to consumer brands too, I feel the story that I hear often is like, we surveyed a couple hundred people in my own anecdotal background and knew there was a white space kind of thing. And there's like this whole other world that you were just mentioning that, okay, you could probably dive a little bit deeper into more research. Sure. Just call it already done research in this space that's more on the academic side to help build a deeper mental model around a space that you're about to go launch a company in. Obviously, the anecdotal aspect is probably the problem you're solving, but understanding how people have thought about the space in the past, I'd say is uber important and very few people do that.

Lilly (08:23): Yeah. And I think with something like a survey questionnaire, be it long or short, there's often an impulse to just throw out a question that you might actually already know the answer to. And you're just sort of there proving your hypothesis, which is important and great. It's good to have data to pat that instinct, but you can also apply something like a desk research audit to the upstream before the questions even go live to enable sort of new developments and questions, new of thinking. And the example I can share is when I was sort of back in communications consulting, we were working with a utility company, a water utility company, that was all about how do we get people? It was in California. How do we get people in California to conserve water more regularly, right? Super important.

Lilly (09:15): From a messaging and comms and creative perspective, it's, oh, the environment, it saved money, this, that, the other, but everything behavioral science has ever taught us is that speaking to the environment or speaking to cost savings isn't actually an effective motivator because you're talking pennies on the dollar and saving the environment is sort of this big challenge ahead. So what we did is we have the opportunity to conduct survey research, but even before we got to that stage, I dove into the desk and went, what is it that really motivates people to save water? And then I can use a really sharp survey question to justify or validate that hypothesis with this specific population.

Lilly (09:58): And so the insight that I found was that people are more likely to do most things, but in this case sort of save water or exhibit what you would call pro environmental behaviors, good environmental behaviors. If you use an appeal to somebody's sense of identity, that kind of virtue signaling of, oh, poor southern California and Californians save water. Oh, of course I'm a Texan. This is a famous campaign. Of course, I'm a Texan and Texans don't litter. That's actually a really famous campaign that really cleaned up the litter in the state of Texas with a message that said, don't mess with my Texas and don't throw your trash out the window. Very effective. So it's appealing to the California identity. Let's test that among our customer base with the survey.

Lilly (10:43): And the question that we asked was as a Californian, so you're sort of setting that context of it is your identity. As a Californian, how important is it for you to save water? I mean, results off the chart, right? In some ways it's a hard question to say no to, and in a perfect world, I might've AB tested it, one where we didn't have California and we weren't working with you all, unfortunately, so we didn't quite have that agility, but we proved the hypothesis. We put the campaign together. So again, that lit search, I think took me like eight hours, which sounds like a lot of time, but actually it had a disproportionately high impact on the campaign and that work and hopefully people's actual behavior is to save water, knock on wood.

Mitch (11:31): Yeah.

Matt (11:32): Yeah, that's awesome. I don't know if it was a water heater, but there was this company that was trying to figure out how to increase sales for, it was something like a water heater or a dishwasher, something that is not like a car that lives in your driveway that kind of has that viral component to it. And they did a bunch of customer research and they landed on this idea of allowing these customers to put a sticker in their window that said they used like an energy efficient X.

Mitch (11:57): Yeah.

Matt (11:58): And like once one neighbor started putting a sticker on, there was this kind of viral effect of, okay, they were able to almost take over neighborhoods. Did we have this conversation, Mitch?

Mitch (12:08): No, not at all, but I wanted to keep nodding, so you'd feel good about it.

Matt (12:13): Okay.

Mitch (12:13): I think that was shown as like one of the reasons why tote bags, which seemed like totally useless things...

Lilly (12:20): Tote-ally useless.

Mitch (12:22): Oh, man.

Lilly (12:22): Heyo.

Mitch (12:24): Can I claim a pun?

Lilly (12:27): It's the people's now.

Mitch (12:28): All right, we'll go halfsies on it, I guess. Why does that seem to work when it seems like they probably already have eight in their closet and it's like, well, now you have a way to say, I use this thing, this product, this service that otherwise you wouldn't know that I use, right? The red campaign, I think is a good example of that too, right?

Matt (12:49): Totally.

Lilly (12:49): Voting.

Mitch (12:50): Yeah.

Lilly (12:51): Voting. You get a sticker.

Mitch (12:52): Exactly.

Matt (12:53): Your comment around like California when it definitely sparks some ideas, but also we do a fair amount of research on our end is like, how do we increase survey completion rates? And some brands have completion rates that are just off the charts, like 80%, 90% plus. And our conclusion, diving into load time, number of options, more or less your standard survey analytics on the technical side, have yet to be able to really prove why that's so high. And I think part of it is just the brand. The consumer is so excited to engage with the brand that they're more likely to then respond to the survey. And it's kind of to your, as a Californian, it's like, what do you feel? It's like, almost as a new customer of X, can you answer this? And it's almost like increased engagement, but it's also allowing them to identify in a particular way.

Mitch (13:44): Yeah.

Matt (13:45): Yeah. The question writing, even for these short things is just so important in extracting meaningful, but also a very high response rate.

Lilly (13:53): Absolutely. And that starts with sort of acknowledging the role that your brand plays or might be about to play in their lives. And I'm thinking carefully about the question in reference to that. The other kind of thing I will say, and I just had this idea, is that even if completion rates are comparatively low relative to other brands, which is almost hard to kind of make that apple and oranges comparison and you wouldn't want to, how else can I leverage that space as an opportunity, right? Or what else can I be doing by just asking that question to get people thinking about my brand differently, right? If you were to go out and say, if you were to die tomorrow, please don't ever ask this of your customer, if you were to die tomorrow, who would you want this purchase to go to?

Mitch (14:39): Interesting. Yeah.

Lilly (14:42): Again, please don't ever put that survey question live.

Mitch (14:47): That sounds like a Sam McNerney question that gets asked I think.

Lilly (14:50): I am lucky enough to say that he was a big part of my early days training and sort of taking the academic in me and evolving that into a functional business professional. It was definitely a Sam question.

Mitch (15:07): Nice.

Lilly (15:07): But in this sort of really terrible example that I've thrown into the ring, it's less about the answer to the question and more about like, it plants an additional nugget, even if you aren't getting folks to respond, which is the other piece of, if you're putting survey questions out there that aren't innovative, that aren't pushing the envelope, you sort of run the risk of one, them assuming that, not that it's spam, but that, oh, I don't know where my data's going. I don't know what it's doing. Do they really care? But two, that's how you're showing up as a brand, right?

Lilly (15:41): You're putting kind of a standard boilerplate question that I've seen before, and I don't get a sense that you care about me, right? If the words don't even mention like the product team behind the shoes you just bought are so excited for you to have them. They would love to know why you purchase these shoes so that they can keep making shoes that you love in the future, which then sort of it feels personal. It feels like your brand is talking to you and it feels like that data is going somewhere.

Matt (16:13): We had a customer ask us last week. I was like, he goes, I want the functionality to just be able to ask 20 random customers a really silly or stupid question. And it was just because like, we have their attention and who knows what might come out of it. Yeah, I want to ask everyone, it's not important, but it's just like, ask someone, was it raining when you placed your order today? Just like something random that probably isn't going to provide a value, but someone might start writing in an open-ended response. Just like, well, actually it was and this is why I thought of your brand because of the rain. So I would have never thought about that. And something that started off maybe as a little gimmicky, like, hey, let's just edit this question and it turns into this whole dialogue internally around like, okay, maybe we should think about this thing.

Lilly (16:58): It's sort of interesting because sort of the, I'm just going to use the word industry, kind of marketing all these folks as a whole have really made this shift towards data-driven thinking and strategy and insight, which is great. We all want that. But the reality is that human brains, they pay more attention to stories and they've actually shown this in kind of FMRI machines and neuroscience that when you're told a story, your brain sort of lights up in all of these exciting ways, versus if you're given a data point. It doesn't necessarily have the same reaction, even though the data point in the story might be communicating the same thing. And so oftentimes, it's one qualitative response out of a thousand that you look at and go, wow.

Mitch (17:43): Yeah.

Matt (17:43): We have a customer who uses our open-ended question functionality, grabs all their responses and then sorts by character length because he's really eager to read what the people who are typing the most have to say.

Mitch (17:58): I feel like we just had this conversation, Matt, about net promoter score, right? And one of the flaws of NPS is that concept of like, well only nines and tens are the real success stories, but that there was this evidence done through some research that showed the six, sevens and eights were the people who were the most knowledgeable of the product or service and they had the most to say.

Mitch (18:20): The fact that if I know a lot about something and I like it a lot, or I like a particular product, and I give it an eight, that really means I like it a lot and in the grand scheme of what it could be, I'm very satisfied. As opposed to someone who knows little about it and is just like, this is a 10 because it's amazing and it's great, but they're not going to be as big of an advocate as someone like me who knows the ins and outs of the industry or of the problem.

Lilly (18:45): Yeah.

Mitch (18:46): What are some other tactics that kind of just pop up in your mind as far as best practices people should be aware of?

Lilly (18:52): Ask yourself, what's the true problem that I'm trying to solve? And oftentimes that can sort of be your steward and guiding light throughout the research and insights process, whatever that may look like. I think I already spoke to this, but there is more data out there than you think. Google is a powerful tool. If you can't collect survey data or long qualitative insights, go to Reddit. Are your customers there? Just spend time in the digital landscape, getting to know your customer through effectively free mechanisms. Do a desk search. It's great.

Mitch (19:25): In reference to that whole NPS thing, that's basically my perception of, well, there's a few perceptions of Reddit, but as far as what we're talking about, Reddit is basically the land of your sevens and eights, right? The people who know a lot, maybe more than you, about what you're doing and have a lot to say about it and that's a gem for market research. Yeah.

Lilly (19:49): 100%. One that I often have to remind myself as sort of an advocate on behalf of a brand is you are a customer. You are not the customer.

Mitch (20:00): Oh, preach.

Lilly (20:01): And so by your instinct to treat everybody as you treat yourself because their reason to engage with your product, it might be different than yours. And finally, and this is my favorite one, is test, test, test. We live in this amazing digital world. There are excellent survey partners out there like nQuire where you can just throw out quick stuff into the universe. And there's really a universe where you can still sort of be meeting your growth goals or communications objectives while still testing things along the way.

Mitch (20:32): Good, good guidance, and a pleasure to have you on to drop that knowledge on us. I appreciate it.

Lilly (20:37): Good, glad to help.

Mitch (20:40): And look forward to see what you do with your new project here.

Lilly (20:43): Thank you for having me. It was wonderful to be here.

Mitch (20:57): That's all for today, folks. Thank you for listening, subscribing and rating the show. You want to learn about Lilly's work to bring the benefits of decentralized finance to smartphones worldwide, check out or follow at @koflerlilly on Twitter. You want to chat with the Fairing team, head on over to See you next time.

 Ready to better know your customers?

See Interactive Demo