Improving Trust To Create Better Health Outcomes: Sandi McCoy and Aarthi Rao

The value of behavioral economics as a strategy to improve health

“We’re trying to leverage our knowledge of human behavior to influence behavior to benefit an individual. And it’s interesting because by the time that I had finished graduate school, a lot of health researchers had lost enthusiasm for the traditional strategies of behavior change. In public health, these are behavior change strategies that rely on filling information gaps or relying on people’s intrinsic motivation to remain healthy. It wasn’t really until the emergence of behavioral economics as a strategy to improve health that there was this renaissance now in the interest in behavioral science again.”

The role of design thinking in public and global health

“And I think elements of [design thinking] have been part of public health and global health for a long time. I think back to some of the work that’s been done over the years on sort of social franchising and social marketing, and a lot of those tools kind of have similar concepts of really empathizing with users and understanding sort of their needs and behavior. But I really think it’s only been in the last couple of years that it’s turned into something more concrete where people have a shared language and they’re sort of more proactively applying it to public health challenges.”

The use of behavioral science for uncovering the underlying problems that public health programs try to solve

“But what might actually be more efficient to look at is what’s the underlying bias or problem that that [a program] is actually trying to solve for. And what’s the literature base for those kinds of solutions. So in some ways, essentially we’re solving for some kind of problem or challenge or barrier or pain point in the process. And there might be very different programs that are trying to do something very similar in a completely different health domain. And so by plugging into behavioral science at this very particular space, it actually helps us in some ways, both narrow and broaden what we need to look at in terms of the literature base.”

Linking behavioral science to design thinking to enhance contraception use

“What [The design thinking process] revealed to us is that we would like to, in some ways, help them think ahead and maybe access contraception before they engage in sexual activity with a partner. They don’t find themselves in that bind. So that’s essentially the insight that was revealed from the design thinking process. But what we then did is map all of these insights against the underlying cognitive biases that we know from behavioral economics… Essentially by doing this sort of infusion of behavioral economics at a key moment in the design thinking process, we’re increasing the likelihood that our ultimate solution going to be effective.”

Leveraging behavioral science and design thinking to achieve universal solutions based on fundamental tendencies

“The design process allows us to identify these commonalities between different segments of people. And if we can distill that down using our tools from behavioral science, it allows us to transport a solution or intervention from one setting to another. And this is the space that I think we would really like to do in the future is for example, take this interesting loyalty program that we’ve developed for girls in this one region of Tanzania. Can we distill that down to the underlying motivational approaches? And could we potentially implement it somewhere else, maybe in different packaging, but again, using those same motivational levers and see if that program works somewhere else.”

• Marrying Empathy and Science to Spread Impact (Article)
• Fostering Behavior Change for Better Health (Article)
• Launch of AmbassADDOrs for Health (Blog article)
• Advancing HIV Prevention and Linkage to Care Among MSM With Gamification (Research publication)
• An Intervention Using Gamification to Increase Human Immunodeficiency Virus and Sexually Transmitted Infection Screening Among Young Men Who Have Sex With Men in California: Rationale and Design of Stick To It (Research publication)
• Pilot study of a multi-pronged intervention using social norms and priming to improve adherence to antiretroviral therapy and retention in care among adults living with HIV in Tanzania (Paper)

• Fostering Behavior Change for Better Health (Article)
• Ten Simple Rules for Cultivating Open Science and Collaborative R&D (Paper)
• Marrying Empathy and Science to Spread Impact (Article)
• Pilot study of a multi-pronged intervention using social norms and priming to improve adherence to antiretroviral therapy and retention in care among adults living with HIV in Tanzania (Paper)

Brooke: Hello everyone, and welcome to the podcast of The Decision Lab. A non-profit think tank dedicated to democratizing behavioral science. We conduct behavioral research and consulting projects with clients such as the Bill & Melinda Gates Foundation to help improve outcomes for all of society. My name is Brooke Struck, research director here at TDL. And I’ll be your host for the discussion.

Brooke: My guests today are Aarthi Rao, director of the design and innovation lab at CVS Health. And Sandi McCoy associate professor in the School of Public Health at the University of California, Berkeley. Aarthi and Sandi, hello.

Sandi: Hi everyone. This is Sandi McCoy and I’m delighted to join the podcast today. So I’m an associate professor of epidemiology and biostatistics, and I study how social economic and cultural forces influence sexual and reproductive health with the ultimate goal of identifying innovative and scalable strategies to overcome global health challenges.

Aarthi: Thank you Brooke for having us on. My name is Aarthi. As you mentioned, I work with CVS Health on design and innovation lab, which is really dedicated towards sort of health behavior changes and finding new ways to get patients to better health outcomes. What I’m really excited to share with you today is the work that I’ve been doing with Sandi over the last couple of years, which has really focused on bringing tools like design thinking, behavioral science, even more traditional product management frameworks into global health.

Brooke: Yeah, that sounds really great. I look forward to diving into that. Before we do however can you tell us a little bit about how the two of you met and how you started working together?

Sandi: Sure. So back in 2014, I was asked to give a guest lecture in an impact evaluation course at the Haas School of Business, focusing on some of my research activities. And Aarthi was actually an MBA student at the time in the class. And we struck up this conversation afterwards and she subsequently invited me to participate in a panel discussion on social impact.

Sandi: And I had these brief discussions with Aarthi, but they were really enjoyable because I learned a lot about her viewpoints on the health-related issues that I was studying. And this was really interesting because she had a very unique perspective. She was pursuing an MBA and had a different lens for many of the global health challenges that we were studying. And around the same time we had a new project that was funded. And I immediately thought of asking Aarthi to join our team and the rest is history.

Brooke: And about how many projects have you collaborated on since then?

Aarthi: I think since we met, it’s probably been about two to three highly specific research projects, but I think what’s been really neat about working with Sandi over the past five years is that we’ve sort of started to expand our definition of kind of like how to use these tools together.

Aarthi: So what started as kind of a number of discrete research projects, largely focused on things like HIV has grown to kind of encompass a lot of thinking we do about these methods and how they might apply to policy and innovation in different ways. Which has been really neat. And I think they’re doing that. We’re also now starting to think a little bit outside of our sector. So thinking beyond academics and beyond the private sector, but also like how does this work for social enterprise or for funders or others. So that work’s been ongoing now for a couple of years.

Brooke: You mentioned exploring this combination of behavioral science and design thinking. To help to put some sort of flesh on the bones of that can you tell us a little bit about how you define behavioral science and how you define design thinking? And then we’ll start to unpack how those two things come together.

Sandi: This is Sandi and I’m trained as an epidemiologist, but we use a lot of behavioral science in our work. So I can give this a stab. In simple terms, we think about behavioral science as the study of human behavior. It’s bringing together this diverse set of methods and insights to understand how and why people behave the way they do. And in our health-related work, we use this knowledge to hopefully influence how people behave and to maybe tip the balance toward engagement and beneficial services or programs. And in health, this might include things like encouraging people to exercise, to stop smoking, or to eat a healthier diet.

Sandi: So we’re trying to leverage our knowledge of behavioral human behavior to influence behavior to benefit an individual. And it’s interesting because of the time that I had finished graduate school, a lot of health researchers have lost enthusiasm for the traditional strategies of behavior change. And in public health, these are behavior change strategies that rely on filling information gaps or relying on people’s intrinsic motivation to remain healthy. And it wasn’t really until the emergence of behavioral economics as a strategy to improve health that there was this renaissance now in the interest in behavioral science again. So now there’s a lot of people who are very excited by engaging in behavior science research.

Brooke: And how about design making? How does that sort of come into the mix? Maybe you can tell us a little bit of sort of a parallel history there about how it emerged, but also how it kind of found its way into this space.

Aarthi: Yeah. This is something that I think Sandi and I talk a lot about. And design thinking can sort of mean different things to different people. But I think from my perspective, I really view it as a process that really helps you creatively problem solve, but most importantly kind of helps you move through ambiguity. Particularly when you’re starting with sort of a blank slate and you don’t have a lot of analogs or maybe baseline data to make a comparison against.

Aarthi: And I think design thinking offers like a really concrete and flexible set of tools that you can bring to that you can sort of get closer and closer to a solution space that makes sense or an intervention design that makes sense. And it’s interesting because I worked a lot in public health before going to business school and getting really exposed to design thinking.

Aarthi: And I think elements of it have been part of public health and global health for a long time. I think back to some of the work that’s been done over the years on sort of social franchising and social marketing, and a lot of those tools kind of have similar concepts of really empathizing with users and understanding sort of their needs and behavior. But I really think it’s only been in the last couple of years that it’s turned into something more concrete where people have a shared language and they’re sort of more proactively applying it to public health challenges.

Brooke: You mentioned public health as a really important area that you’ve been applying this combination of behavioral science and design thinking. Could you walk us through kind of the life cycle of one of those projects? Along the way would you mind just sort of calling out or flagging where you see specific elements influenced by behavioral science or design thinking during that journey?

Aarthi: Yeah, definitely. And I think kind of before describing how they’re using combination, a useful starting point is what often happens when you’re not sort of using behavioral science or design thinking. And I think a lot of times in the public health space, there’ll be a top down priority. That’s largely dictated by like a biomedical need, right? You have a need for a vaccine or a need for a drug or a diagnostic to sort of solve a specific problem.

Once the R&D on that is done, it then kind of like goes out into implementation and lots of folks take different approaches. But it’s typically a couple of funders, or a couple of governments kind of making big choices about what that looks like. And so I think where design thinking and behavioral science really come in is once you get past that sort of biomedical R&D phase, and you’re really starting to think about how do people actually access and use and benefit from these commodities.

Aarthi: And so when I think about the life cycle, I think it really starts with really understanding users. So it’s interesting to me, Sandi mentioned implementation science earlier. I think design thinking and implementation science sort of have some parallel paths that come together really nicely.

That really start with one, just understanding who are you trying to design for? Who’s that person that’s going to take the treatment and use that diagnostic and understanding sort of in their own words and in their own perspective, like what benefit it’s going to have for them, but also what are their sort of contextual barriers that might get in the way.

And so really mapping that out, which in design thinking lingo is kind of the like empathize phase. And then using sort of those open-ended insights about users and kind of defining those into more concrete needed statements.

Aarthi: And this is a point where I think behavioral science and design thinking diverge a little bit. And that often design thinking is really describing needs in terms of sort of the user’s perspective. And this is one stage where I think if you’re thoughtful about it, you can kind of bring the disciplines together. So defining a need user might care a lot about privacy or about some of their social goals. But then sort of you can strengthen that with evidence about human behavior and kind of what we know about biases.

Aarthi: And so I think once you do that stuff, you’re in a much better place to then go into prototyping. So this is where you start to get a lot of real world feedback. And I think what’s really neat about this stage is you can take a lot of different ambiguous solutions and kind of create low fidelity sort of prototypes that let you get feedback really quickly and move you again just a little bit closer to that right solution space.

Aarthi: You might not be at a point where you’re making big investments and kind of big choices, but you’re kind of taking a little bit of risk out of the equation before then kind of going into that more pilot, evaluate, launch phase.

And I think in those latter steps is again where behavioral science becomes really useful because you can start to implement some of those nudges or really think about like choice architecture in your design of prototypes or in your design of products before making really big investments to scale them. Does that kind of make sense?

Brooke: Yeah, I think so. Sandi I do invite you to chime in as well on this.

Sandi: Yeah, I think I can maybe give an example for one of the ways that we’ve done this. So we were thinking that there are, for example, are opportunities for behavioral science to be infused into this design thinking process and mindset to really strengthen the prototypes and ultimately results in a more effective final solution.

Sandi: So one of the projects that Aarthi and I have been working on is a project in Tanzania to help increase young women’s access to contraception and HIV self-testing at private drug shops. So one of the insights that resulted from our design thinking process was that young women are reactive with respect to contraception.

So they’re often able to obtain emergency contraceptives after the fact, but they’re not really thinking ahead about planning for the future. They’re essentially just taking action for acute needs when they’ve had sexual encounter and then they need to make sure that they don’t then become pregnant.

Sandi: And so what that revealed to us is that we would like to, in some ways, help them think ahead and maybe access contraception before they engage in sexual activity with a partner. They don’t find themselves in that bind. So that’s essentially the insight that was revealed from the design thinking process. But what we then did is map all of these insights against the underlying cognitive biases that we know from behavioral economics.

Sandi: So the insight that I just described to you is essentially a great example of time and consistency and present biases. And by knowing that it kind of unlocks some opportunities for us to use a corresponding nudge strategy to help change the way that a young woman might view that decision.

So we might try to nudge her to take action earlier by trying to get her to visit the drug shop more frequently and access free contraception and making it easier for her to obtain those HIV self-test kits by giving her one when she first comes to the shop and enrolls in a new program that we designed.

And so essentially by doing this sort of infusion of behavioral economics at a key moment in the design thinking process, we’re increasing the likelihood that our ultimate solution going to be effective.

Brooke: Right. So that’s good. I think there’s something really tangible and clear there that our listeners can take away from this. It’s that’s often in the design thinking process, one of the steps where perhaps some people feel that there’s not as much guidance, which sometimes can be beneficial because it leaves you a lot of latitude, but other times it can be a bit challenging because it doesn’t offer you much direction is how you come up with the interventions that you’re going to test in those early pilot phases.

Brooke: So if I understand you correctly, behavioral science is supplying you at least sort of one taxonomy that you can work with and thinking about what kinds of interventions or what kinds of targets for interventions might be worth going after. Is that right?

Sandi: Yes, that’s exactly right. Actually it helps provide some directionality to what the prototypes might look like or what the most likely prototypes to succeed might look like.

Sandi: One other thing that I think actually is very valuable as part of this intersection of these two approaches is that at least in public health where I work, there’s this very deep evidence, scientific evidence base for many of these strategies.

And so if you can figure out what is the underlying issue of each insight, it then helps you sort of narrow down the literature that you would need to review, but also helps prevent you from reinventing the wheel. If you essentially think that maybe the answer to a particular problem is maybe we need some peer navigation or a peer support group.

There might be a very deep literature on that where people have tried it and it either was successful or failed. And this would actually help you conduct a more efficient literature search of the scientific and literature base, which I think is very valuable as part of this process as well.

Brooke: Yeah. So that’s interesting. There’s definitely a component of behavioral science that draws from this tradition of doing extensive literature review and sort of background canvassing before you start articulating a new hypothesis.

Because, as you rightly pointed out, you don’t want to be reinventing the wheel, you want to be building a body of knowledge. In design thinking do you feel that that’s perhaps something that’s a bit lacking that there’s less of an emphasis or less of a focus on canvassing the background of what’s already out there before you start developing candidates solutions?

Aarthi:  Yeah, I think it’s definitely not an inherent part of the design thinking process to like do that extensive evidence review. And that’s something that we find like just works really well. Sort of being open ended up front, sort of taking that beginner’s lens to understand what are all the facets of the problem. But to Sandi’s point, once you start to narrow in on which aspects of that problem are most important for the particular outcome you’re interested in, that’s when it’s incredibly useful to make sure that you’re sort of standing on top of the foundation of everything that’s come before you.

Aarthi: I would caveat that with it’s a nice balance, but you don’t want to run into the evidence too quickly either. So I think one thing that I really appreciate about the design thinking influence is that you are taking that time to be open and make sure that you’re going after the right question and the right problem. Because sometimes you see the flip side where someone might be really committed to a social norm solution or a pledge solution. You don’t want to be sort of a tool looking for a problem. Whereas design thinking, I think really pushes our team to think about sort of the problem and all of its facets fast.

Brooke: Yeah. So there are these two phases in the early parts of a design project, right? The ideation phase, where things are supposed to be really wide open and unconstrained. And then this second phase pruning where you start to prioritize which intervention points and perhaps which intervention types you think are the best candidates to move forward eventually.

Brooke: So if I understand you correctly, the idea is to keep the blue sky very blue when it comes to the ideation phase. You really want to have an open mind and generate lots, lots of different ideas before you start to narrow down in that pruning phase. Is that correct?

Aarthi: Definitely. And I think it’s less about sort of a one-way funnel that gets narrower and narrower, but creating the right feedback loops. So even if you have a lot of ideas, I think you don’t want to cut them off too soon. So really a great example of this is from the project Sandi was just describing where we’ve been working with drug shops in Tanzania.

And we are trying to figure out like what’s the best way to help girls get access to HIV self-tests? And so we did a lot of work, spent time with drug shop owners, with girls, with their families, really understanding their day to day lives and came up with like five or six, very, very different solutions that were really going to tackle different aspects of that problem. And so we could have added in the behavioral science like right away.

Aarthi: And we did to some extent in terms of the insights behind the ideas, but the first thing we did was actually just go out and share these ideas in the form of storyboards or other level of fidelity prototypes with girls. And we were quickly able to kill two or three of them.

And then when we started to converge on the thing that made the most sense, we could get much more specific about drawing in the evidence and drawing in some of the past work that’s been done.

Sandi: If I can just add one thing that I think is really cool about the process that we’ve been using is that in some ways by starting to pull in the scientific literature from behavioral science, more sort of midway through the process as Aarthi’s describing, it actually allows us to understand that we need to look in the literature, not just for this specific problem or the specific kind of approach that we might be investigating.

Sandi: For example, one of our prototype ideas was around a delivery program. So maybe girls could call a hotline and have contraception or HIV self-test gets delivered to them or at a place that they chose. And we were excited about this idea and it actually subsequently did not test very well among the girls that were part of our research studies.

So we killed that idea. But we might inadvertently think, oh, let’s look through the literature to see that what other kinds of delivery programs for contraception have been out there? Has anybody tested these before?

Sandi: But what might actually be more efficient to look at is what’s the underlying bias or problem that that delivery program is actually trying to solve for. And what’s the literature based for those kinds of solutions. So in some ways, essentially we’re solving for some kind of problem or challenge or barrier or pain point in the process.

And there might be very different programs that are trying to do something very similar in a completely different health domain. And so by plugging into behavioral science at this very particular space, it actually helps us in some ways, both narrow and broaden what we need to look at in terms of the literature base. Hopefully that makes sense.

Brooke: Yeah. I’m really starting to see the benefits that you’re talking about. The kinds of rich combinations that are available to bring design thinking and behavioral science together.

Brooke: It sounds also in your descriptions, like sort of in the margins here and there, there are also potentially some tensions, there are some challenges. So I’d like to talk through those with you as well. Because I think will be really helpful to understand why this isn’t just an easy thing to do and kind of a slam dunk that you should just put on your to do list for first thing Monday morning.

So what are the kinds of risks or tensions or challenges that people should be aware of? This is something that they want to explore further. So one that comes to mind from what you’ve said so far is the possibility that behavioral science could impose too much constraints on the design process.

Aarthi: When using these tools, like I’ve never felt sort of super constrained to use one discipline or the other, but I’ve actually found that they flow together pretty well. But I think when there are attentions, it’s typically about sort of two things. So one is sometimes we’re talking about inherently different things. So behavioral scientists, when they talk about innovation they’re sometimes referencing what I call like little I innovation or continuous improvement, right?

Something like a 401(k) already exists or a program already exists. And you’re trying to increase engagement, increase sort of some highly specific metric where it’s actually much simpler to kind of look at behavioral tools, pick the right one, apply it right away and like rigorously measure the difference from baseline. But when sort of like design thinking folks are talking about innovation they’re typically talking about the bigger I innovation. So more transformative breakthrough things where there’s not an analog.

Aarthi: So you’ll have to sort of invest much more time upfront to figure out what is that opportunity that you’re going after. So I think about examples like Blockbuster versus Netflix is a great example. You can sort of look at Blockbuster’s old business model, apply behavioral science to a lot of levers in that model and likely have increased engagement adoption, but it wouldn’t have necessarily saved their business. Right.

But something like a Netflix that fundamentally change, like how are people interacting with entertainment? How are they experiencing something? Required understanding people’s preferences and unspoken sometimes completely unconscious desires and turning that into something very tangible. So I think that’s sort of the biggest check box that you have to consider is, are we talking about sort of the same level of ambiguity, the same maturity of problems, and if you are, then I actually think these tools work really well together.

Aarthi: And I think as long as you have the right team, it comes down to sort of the people and the collaboration. I think Sandi alluded to this earlier, but one of the reasons we started working together is just from the sort of intellectual curiosity of wanting. I was super interested in behavioral science and how that could work for public health and business. She was really open to sort of new approaches and trying new types of experiments. As long as you sort of have that level of trusting collaboration. I think there’s not as much tension as it may seem. Sandi would you agree with that?

Sandi: I would completely agree with that. I think we were both in a moment of just openness to creativity and different ways of tackling some of these intractable health challenges that traditional behavior change was not working for as well anymore.

Sandi: And so I think that was just the right spark to allow this sort of fruitful collaboration and process to emerge. I can build on what Aarthi said about thinking about the concept maturity as a process, where do you plug in these different tools using some examples from our own work in health? So one of our first projects together was looking at the problem of HIV treatment adherence. And many of your listeners might know that when people are living with HIV infection that starting HIV treatment as early as possible is incredibly beneficial, both for the individual, but it also helps to virtually stop onward transmission.

Sandi: So it’s both a strategy to extend people’s length and quality of life, but it also is one of our main tools in ending epidemic spread of HIV. One of the challenges is that many people have a really difficult time being adherence to their medication, particularly in incredibly impoverished settings, where people have to make choices between maybe food or work and going to the clinic to pick up their HIV treatment.

Sandi: And this is especially hard if people are feeling healthy, it’s very difficult to be motivated to taking a daily medication for the rest of their life. And at the time we were working on this problem, there’s actually been quite a bit of work focused on HIV treatment adherence. So it’s standard of care for people to receive HIV treatment adherence counseling when they start treatment. They receive refresher counseling when they have lapses and adherence. And some clinics are using SMS messaging or peer groups to help bolster adherence among people who are struggling.

Sandi: And yet, despite all of these approaches, this is still one of the main challenges we have in making sure that people can live long and healthy lives. And that we can essentially help to end the HIV epidemic. And so Aarthi and I came at this problem from a really different perspective. We recognize that there had been a lot of work in this space. There were already a lot of interventions happening and that maybe we could use behavioral science to help to improve what was already being done.

Sandi: Maybe we should improve the clinic experience, make it a little bit more fun or help to enforce the knowledge that a lot of people are actually being adherent. That that’s indeed the normative behavior. And so recognizing that this was a concept at a farther point in maturity, we actually use behavioral science to develop a really innovative intervention that uses priming and social norms to help nudge people to be more adherent to their medication, come to the clinic on time, and actually have a little bit of fun at the clinic.

Sandi: And in a small experiment that we ran, we actually found that people who were exposed to that intervention were more likely to be adherent at six months compared to people who were not exposed. Now, in contrast to that project, another project that we’ve already talked about today is this one focusing on increasing young women’s access to contraception and HIV self-testing. What’s different about this project is that HIV self-testing is brand new in some parts of Tanzania.

Sandi: In addition, there are currently no programs to help young women access these products at drug shops. So essentially we were starting from scratch and we wanted to think big. We wanted to have these incredible array of blue sky solutions from which we could choose. And the sky was the limit in terms of what these kinds of programs could look like.

And so for that project, which was very early in terms of our thinking, we used the full design thinking process infused with these behavioral science touchpoints, along with another strategy that we often use called implementation science, which is a way to think about the context in which any solution will eventually need to live and thrive.

Sandi: And so by using these tools together, we came up with this new solution that was honestly something we would have never predicted at the beginning of the project.

Brooke: So I’m fascinated. I want to hear more. It’s not like you’re going into this context where there’s already this network of drugstore, drug shops. There are the Blockbusters, and you want to come in and develop the Netflix for this problem. What does Netflix look like for HIV self-testing in Tanzania?

Aarthi: It’s a good question. So how we approach this is what could we really do to transform girls’ experiences of visiting these shops? So what we’ve found is a lot of girls are going to the shops all the time. So on the surface, that’s a good thing.

That means if we put the tests in the stores, they should have access to them. They should be able to use them. It makes it really easy, but in practice, they’re doing really quick in and out trips. They’re not doing a lot of product discovery. And I think most importantly, we found that their day to day lives are just very policed.

Aarthi: So there’s kind of rigid schedules that they’re following, a lot of oversight from parents, neighbors, elder siblings in terms of like what products they’re buying, how they’re spending their time. And most interestingly an important for this project is even the shopkeepers who don’t know the girls well will sort of fall into this gatekeeper role where they feel like they should enforce sort of these community standards that girls shouldn’t be asking about HIV or accessing contraception.

Aarthi: And so a lot of our ideation and thinking came down to how do we sort of disrupt that relationship. And how do we open it up so that girls felt comfortable accessing these product in these shops. And that shop owners are willing to engage. And what’s interesting is from the shop owners perspective, this is a great example of something that is sort of predictably irrational, right? So if they’re selling tests and having customers in the store, it’s actually should be driving up their profits and margins.

Aarthi: But this perception that they shouldn’t be a gatekeeper is so strong that they hold themselves back from doing that. So we just spent a lot of time ideating against this, trying different things. The solution that we landed on was kind of two-pronged. So one was to encourage loyalty. So loyalty cards or loyalty programs are not sort of a new thing. Lots of companies in the US and across the world employ them. But in this setting in Tanzania, it was pretty novel. No one had sort of rewarded a customer for coming back to the same shop again and again.

Aarthi: And so that inherently starts to create some trust between the girls and the shopkeepers. We also just did a lot of interesting things that sort of created a perception of privacy and relationship building. I think the linchpin of this intervention was the actual loyalty card on the back, had a series of code and symbols that girls could just kind of point to.

And it would mean if they pointed to one symbol, it might mean they want a pregnancy test. If they pointed to something else, it might mean they want an HIV self-test. A parent or to an onlooker it’s not necessarily obvious what those symbols mean.

Aarthi: And so it almost created this shared secret between the shopkeeper and the girls. And once we introduced that, it also kind of gave the shopkeeper implicit permission to counsel girls on these products. To offer girl these products. And it’s worked really well. I know Sandi, the team has been kind of tracking feedback from both girls and shop owners. And I think it’s been resoundingly positive. Is that right?

Sandi: It’s been very positive and we’re actually doing a randomized control trial right now to actually measure the effectiveness of it. What’s cool is that we’ve taken the final intervention and mapped it against these biases that are well known from behavioral economics.

Sandi: And so you can imagine that a loyalty program is leveraging things like commitments and incentives, and there’s an element of using social norms that are telling the girls that it’s okay to ask for these products. There’s an element of using salience so we’re using these visual displays where girls come into the shop and they can see, “Oh, these are products that I can touch and feel and ask for.”

And so baked into the solution that has this sort of pretty fun packaging, that’s very appealing to young women is a lot of behavioral science and many different motivational levers that appeal to different people. So that’s another important part of our process is that rather than sort of one nudge, one problem solutions, we’re using a lot of different elements of behavioral science in the same program to hopefully capture many different segments of young women to encourage them to use these products and services.

Aarthi: And I think just to build on that, that comes from sort of taking that customer experience or design thinking lens. Because we weren’t sort of narrowly focused on just that in shop experience, but we really cared about from beginning to end. What would motivate a girl to come to the shop? What would that experience be like once they got home? Could they safely use the product and sort of know what to do?

So we really were thinking about that entire experience in journey from both the shopkeeper side and the girls side. And that helped us sort of identify these moments that matter and pull in different tools and principles where it made sense to do.

Sandi: If I can just add one of the things that I think has contributed to the magic of these approaches has been having a really engaged local team. So these examples we’ve been talking about so far are based in Tanzania and we have an incredible team of local scientists and designers in Tanzania who are playing a major role in these activities.

It’s been very interesting for us because we’ve been doing sort of brainstorms on both sides of the globe and then we get together and we’ll do them together and distilling insights that are coming from many different perspectives. But really a key part of this has been having a diverse team incredibly based locally, and also a team that’s really open to different disciplines. And there’s sort of an inherent value and respect for different disciplines and bringing them in at different phases of the process.

Brooke: I think that that’s a really important point to start pulling open. And I’d like to sort of sound you out on that. It sounds like there’s a combination going on of inherently local stuff that you shouldn’t be able, or you shouldn’t expect to be able to find without feet on the ground. And that’s brought together with some components from behavioral science that are thinking about bias news on heuristics that are supposed to be lodged deep within the evolutionary history of the human brain. And therefore they’re supposed to be much more universal.

Brooke: What are the characteristics of a team that allow them to bring together what seem on the surface like potentially kind of contradictory or conflicting ideologies or views about the world that for instance an ethnographic approach to mapping out the experience of individuals is inherently local. There’s no pretense that what is being mapped out through that kind of ethnography is going to be universally visible.

Brooke: Whereas on the other side, cognitive science, which takes this view from the outside in is really trying to look for those things that are invariant across populations across through time. That kind of thing.

There’s also an intellectual heritage question there that design thinking from what I can see is much more established in the world of professional practice, whereas behavioral science is much more established in the academic realm. How do you find team members that can bring together that locality with a universality background and professional context versus an academic context? Which features help those teams to collaborate effectively?

Aarthi: From my perspective, the most important things are finding folks who have that intellectual curiosity and willingness to learn all of the disciplines. Because I don’t think there’s anyone that comes to the table that’s an expert interviewer, expert behavioral scientist, who also knows how to run the very complicated logistics of a randomized experiments and sort of enforcing data quality in the field.

Aarthi: So I think for our team, we’ve done a lot of hands on training and Sandi jump in if I’m wrong. But I think what we looked for in the beginning was like someone that really cares about the problems is kind of willing to learn these different tools and approaches. But most importantly, a lot of them have stuck it out across projects and that’s been sort of invaluable. So a lot of the same folks that were involved in some of that original work and kind of creating that social norms, intervention HIV clinics are also now involved in the project with the drug shops.

Aarthi: And so having that institutional knowledge across projects has been really useful. But I remember when we were recruiting early on, it was a pretty funny experience kind of thinking through someone who’s interested in design thinking in rural Tanzania when you know that no one knows what that word means.

So we were looking for folks that had interest in journalism and like listening to stories and who are good at extracting stories from people, or looking at folks with a marketing background that we’re interested in sort of driving engagement on different programs and products. And then we sort of had to think through how to bring that all together.

Sandi: We have a really incredible team at this point for the drug shop project that we’ve been mentioning. We have the two of us, we have a behavioral economist, who’s our colleague, Jenny Liu at UCSF. We have a medical epidemiologist, Dr. Prosper Njau who’s in-country. And then we have the local team who is the ones on the ground collecting data and helping to interpret the data.

Sandi: And what’s cool is that just like in the process of the sort of design plus behavioral science process, all of us are stepping up to have a higher level of engagement at various parts of the process. So thinking about as an epidemiologist, I think a lot about data and how are we going to measure impact.

And we think a lot about implementation. And so there are various stages in the process where I can infuse the process with that knowledge, but clearly Aarthi, for example, is heavily involved in the design process and helping us unpack all of the information and stories and narratives that we’re hearing from the field.

And clearly the local team is the most important in terms of interpretation. So they’re the ones are really helping us understand what certain things mean and what are the contextual factors influencing people’s experiences and stories. And so I think just this openness to working together and to understanding that everyone has a role to play has been really important

Brooke: Just from listening to the two of you. It sounds like strong leadership is another ingredient that the two of you are just too humble to mention. So I’ll mention it on your behalf. But one of the things that came to mind listening to you speak is about facilitation. You mentioned that you give training to your group and you actively work on the process of how to do this well on the dynamics within the team to ensure that they are able to execute that. Do you have someone on the team whose role is specifically that kind of facilitation and moderation between different actors within your team?

Sandi: In general, or a project like this in addition to the folks I just mentioned, we would have both a project coordinator type individual that would often be based on my team at UC Berkeley, but it doesn’t have to be that way. And then we would also have a local coordinator as well. So those two would be counterpoints who would be sort of running the day to day operations. And then folks like Aarthi and myself and Dr. Liu and Dr. Njau would be helping to sort of drive the strategic direction of the project and just making sure that we’re staying on track and achieving our goals.

Sandi: The part that’s different from many other typical projects that we might work on is that it requires this level of openness at the outset. So everyone has to buy into the process and agree to sort of follow the process where it leads.

Sandi: And that’s very different than traditional epidemiology or traditional health related research, where it might be very protocol driven, you may have an idea of what you’re going to be doing or what the solution’s going to look like.

Sandi: We often have to write ethical review board applications and make sure that our projects meet the standards set by for ethical human research. And all of that has to be laid out at the outset. And so for a project like this, it looks very different. The solutions aren’t pre-specified, and everyone has to share the vision for being open to where the process leads. For all of us we’re so excited about what comes out of these projects and how much they surprise us.

Sandi: And they’re so different than what we thought we might do, that all of us are really bought into doing this more and more.

Aarthi: Yeah, I think it just takes an incredible level of trust. I’m sure there’s moments in the process where Sandi’s probably wondered to herself, “What is Aarthi talking about that sounds crazy.” But it’s just sort of like that willingness to try it and then see where it goes.

And once you do it a couple of times, and you have different results than you’re expecting or better results than you’re expecting, I think that trust just grows and that lets you then try even more applications on even harder problems.

Brooke: What about incentives? I mean, it’s great to have the right culture within the team to keep it together. And you mentioned this open-mindedness at the outset, but if you’ve got an inter sectoral team, some people coming from academia, some people may be coming from the for-profit sector, the non-profit sector, government.

If you have this kind of open-minded attitude at the beginning about what kinds of solutions, you might develop. Along with that, I feel there’s also kind of the openness about what kinds of outputs this might produce.

Brooke: How do you create a situation where an academic person has an opportunity to contribute in this project fully if it’s not clear that there might be some kind of scientific publication that comes up the other end. And like it or not, this person’s professional events might depend on those kinds of outputs coming from the project. Do you feel that there is a way to sort of have your cake and eat it too? To leave the project open ended enough at the outset to find a solution that’s really going to have impact.

That’s going to help the people that ultimately you’re aiming to help while at the same time, giving it enough structure that the kinds of outputs that are required from the project in order to create a space for everyone to contribute are really on the agenda and they’re not going to fall by the wayside along the way.

Sandi: It’s a great question. I can try and tackle that one. So I’m an academic and so many of the things that you just mentioned speak directly to the climate of academia and the incentives that I have in terms of publishing and scientific productivity.

Sandi: I think one way that I think about this is that the way that I think about my own research portfolio is that I’m very interested in tackling the research problem, the health problem, and I’m not tied to a specific methodology for doing so. I think Esther Duflo has this great quote where she says, I never met an evaluation that I didn’t like.

And I feel the same way. So as long as the solution that we generate can then be rigorously evaluated using these very classical tools from impact evaluation and epidemiology, that really helps me meet my obligations as a scientific researcher, because I can then publish those in peer reviewed scientific journals. And of course with the whole team as collaborators and that keeps me and my superiors happy in terms of my scientific output.

Sandi: But I think what’s critical there is that my research isn’t tied to a specific approach. I think that would be much more challenging if, for example, I was a world’s expert on commitment devices, for example, and I can only study commitment devices. I think that by framing my work around the centrality of the research question as a driver of what we do, that we’re able to really pull in the right strategy for the right problem. And that helps me meet those obligations that come with academia.

Sandi: One other perspective I can mention is certainly our team on the ground does feel some pressure to implement quickly and to show results quickly so that they continue to enjoy great relationships with the government and other local partners. And so as much as possible, we try to make sure that we have built in milestones to make sure that they can also show progress to make sure that we always have great relationships with governments and other local partners. Aarthi I’m sure you have some to add from your perspective as well.

Aarthi: You said it well when you sort of mentioned that sense of being mission driven and committed to the problem. I think that’s probably the most critical piece. If you have a group of folks who care more about the problem and the outcome you’re trying to drive than the exact intervention or process you’re going to take to get there I think that’s when you have a lot of room for creativity and a lot of room to do really neat interdisciplinary work.

Aarthi: For me when I think about like incentives for doing this work, it’s all about solving interesting problems in new ways. And so many of these tools sort of exist in the private sector, but the potential application for both public health, but other policy and areas of social impact is so high. That for me it’s just really interesting to think about how do you apply these tools and what are sort of interesting outcomes and interesting programs you can get to that you might not otherwise be able to achieve.

Sandi: I can say that’s definitely been one of the areas where if there have been any struggles, it’s actually probably been more on my side because as quantitative researcher it’s sometimes hard to sort of unlock creativity and be willing to move into a space where there’s of unknowns and to think about these blue sky solutions without sort of killing them quickly because of all of the operational constraints you can imagine. And I think that’s one thing I’ve really valued from this collaboration with Aarthi is just infused in much more of my work in life than I thought would originally happen, which is just to embrace that creative side of my brain.

Sandi: And to use that to make the work better, more effective, potentially more scalable, maybe more sustainable. And that’s been just an incredible gift I think too much of the work that I’m doing.

Aarthi: Definitely. And I’ve started to use experiments and other areas of my work. And it’s one of those things that once you start doing experiments and getting used to seeing results in that way, it’s sort of impossible. So it definitely the work that we’ve done together is starting to have I think bigger impact in unexpected ways.

Brooke: So from what I’m hearing, this approach is something that’s not along that clearest, straightest, most obvious path for professional advancement. There are challenges that are there, but also on the other side of the equation, they’re not insurmountable. It can be done.

And from what I’m hearing from the two of you and the fact that you’re here talking to me together today about the projects that you’ve worked on, it’s a testament to the power of those problems that are so worth solving to motivate individuals such as yourself to push through those potential barriers, to not worry so much about the uncertainty professionally that might be somewhat higher by taking the path less traveled or things like that. Does that sound about right?

Sandi: I think that’s right. I mean, the way you framed it, it sounds like Aarthi and I are major risk takers. And I don’t know, I would say that, but one thing I think that has come up for me is just that it being willing to go on this journey, is thinking a lot about my role as a scientist.

And what does this mean for a science? And one of some of the interesting feedback I’ve received is questions about something you mentioned early about whether these kinds of approaches are only relevant in local spaces, or is there any kind of universal generalized ability that we can say about these kinds of solutions? And this is something I think Aarthi and I would really like to do in the future. The essence of this question is this question around generalizability. So as an epidemiologist, we always wonder can the findings of a specific research study apply to some external population.

Sandi: And because design thinking is often just inherently local, one of the challenges that I face in academic spaces is this question around generalizability. And I think about it in a couple of ways.

One is that we might have achieved the gains in public health that we’re going to achieve with big structural policy or programs. For example, if we think about the problem of HIV treatment adherence, we can fix the supply chain, we can remove financial obstacles. But once those strategies are implemented, we’re really talking about an individual level behavior change.

And so a local approach might be what’s needed. So it might be the case that we need to see more of these truly multidisciplinary approaches to get us to the last mile of our goals, because there’s no longer these big easier structural challenges that we can do to from a top down. We’re going to have to take a more bottom up approach.

Sandi: I think another element of this is that the design process allows us to identify these commonalities between different segments of people. And if we can distill that down using our tools from behavioral science, it allows us to transport a solution or intervention from one setting to another. And this is the space that I think we would really like to do in the future is for example, take this interesting loyalty program that we’ve developed for girls in this one region of Tanzania. Can we distill that down to the underlying motivational approaches? And could we potentially implement it somewhere else, maybe in different packaging, but again, using those same motivational levers and see if that program works somewhere else.

Sandi: And so it’s not like taking the exact packaging, the exact same intervention, and then just plopping it down somewhere else. No, we know that’s not going to work given the specificity of the solution, but it’s the underlying design principles that might be transportable. And so this is a really interesting way to start to think about programs and policies is actually viewing them through this behavioral science lens and using that to help influence where they can be transported to.

Brooke: Yeah. So for instance, if we think about that loyalty card, it’s not necessarily the card itself, it’s kind of this symbolic functioning of that card that by creating a more discreet or a less obvious form or kind of manner of communication, one that’s less prone to eavesdropping for instance you can create social license for certain kinds of behaviors that otherwise wouldn’t be licensed.

Brooke: So if we look at a similar problem in a different context, it’s not necessarily the case that a loyalty card with this symbolic coding on the back is what we’re after. What we’re after is that functional thing that the loyalty card fulfills, which is creating opportunities for social license. And doing that specifically in this case, through the creation of a symbolic language that is kind of hard to eavesdrop on. That actually brings me around in my reflections to a problem that people have pointed out with behavioral science in past is that some nudges lack durability.

Brooke: So you roll them out and when they first arrive on the scene, they’re very, very effective, but then over time, people become kind of desensitized to them and the effects kind of wear off. Do you anticipate that something like that might happen with something like this loyalty card program, where right now it’s a discrete form of communication between the girl who’s coming into the shop and the shopkeeper. But as that symbolic language, this discreet mode of communication between them becomes better known it kind of loses its potential because all of a sudden it is easier to decode the script if you will?

Aarthi: I share that concern sometimes if you use nudges again and again they become less effective. And I think that’s why the design thinking toolkit is so useful because you view it in the lens of that entire customer or patient or user’s experience. And so you’re not just going to focus on that one interaction. So for us the loyalty program, although we went into this project really focused on HIV care, it actually addresses a number of products and services that girls go to these shops for.

So it’s about mapping that entire experience and finding something that’s going to work because we couldn’t have created something just around HIV or just around contraception. It sort of had to fit within the shopping experience. But because we’ve done that and because we’ve mapped it and we really understand both the shopkeeper and girl’s perspective, it allows you to then iterate and pivot as needed.

Aarthi: You could evolve that program. Let’s say a few years from now with mobile phone penetration increases, you could evolve that to a text based program or a smartphone based program. But the foundational insights of this gatekeeper role that the shopkeepers fall into, or the level of privacy that girls expect, those won’t change. And those can kind of be your really foundational design principles that you apply to any intervention.

Sandi: That’s exactly right. And I think that that’s a trap that academics in particular can fall into is that it’s straightforward to think about, oh, I have an interesting principle from behavioral economics. I want to test it. This sort of one nudge for example, and you can show in a nice, beautiful, rigorous experiment that, oh, that works in the short term, but is it going to work in the long term?

Sandi: And so to build on what Aarthi said, we’re hoping that we’re designing comprehensive solutions that are including multiple strategies, so that if one sort of loses its novelty quickly, there are many others that are included in the solution. But also just to sort of underlying something she said around the value of the design process. In a typical project, you again, might be sort of approaching this from a one barrier one solution approach. Whereas for this specific project and the other ones where we’ve used this approach, we have a comprehensive list of insights to which we can go back to at any point and say, let’s come up with another strategy that’s going to address the underlying bias of this insight.

Sandi: And how can we address this to do this in a fresher way. If we need to give girls the implicit permission to ask for these products, there are multiple ways we can signal that. And so again, it’s really highlighting the benefits of using the two approaches together because the design process is doing this incredible contextual mapping and landscaping for us to understand what’s actually happening. And we’re using the behavioral science to put some more rigor behind the approach, but we can continue to iterate over time with this sort of incredible richness of information.

Aarthi: And I think the funny part about that is it’s not actually super innovative, right? That’s sort of like basic business 101, you expect the market preferences to change. You expect market forces to change. So when a company puts a product or service out into the market, they’re immediately planning for like, what’s the next version? What’s the version after that? Because they know the market’s not static and preferences are not static. So I think that’s a really valuable piece about design thinking. It’s sort of like expecting that to happen and gives you a set of tools and ways to think about that so you’re sort of prepared for it.

Brooke: Shifting the conversation from this binary question, will the intervention last, will it not to more of a scaler question to say, okay, well the intervention is going to have a certain durability, a certain half-life to it. How long is that do we think, and what are we doing now? And what’s our plan for the future to make sure that we’re taking account of that rather than just saying, well, this is a thing that’s a one and done. We don’t need to worry about it. Or this is the thing that is eventually going to lapse. And so we need to sort of start building something already.

Sandi: Yeah, that’s right. One of the things I’ve actually been frustrated with in the health field right now is that many creative ideas and solutions to some of the global health challenges are sort of shot down very early in the process because they’re not perceived to be sustainable.

Sandi: And I actually think that we need to give a little bit more space to develop sort of early stage innovation and developing a sustainability strategy should be part of that process. And as Aarthi said, of course, these are ideas coming from the private sector that are completely normative activities. And they’re new to public health because we do have a bit of a sort of black and white viewpoint where it’s either effective or not effective. And it’s either sustainable or not. Whereas if we took a lens, that’s a little bit more like continuous quality improvement.

Sandi: If we think through that lens, it actually makes a lot more sense to be using behavioral science, to make these sort of small, incremental improvements on projects that maybe had their Genesis in a design thinking process. Using the right strategy at the right moment of a concept maturity can move us more from this sort of dichotomy of effective or not, sustainable or not, towards how do we take all of the knowledge that we have around certain approaches in health and just make sure that they stay effective and stay improving over time as the populations that we’re trying to reach changes over time?

Brooke: What is the best case scenario for the future of combining design thinking with behavioral science? What do you think are the most exciting application areas? What’s the biggest thing out there? What’s the big light at the end of the tunnel?

Aarthi: I think for me, what’s exciting is that we’ve talked a lot about sort of public health and healthcare applications. But honestly I think this combined framework could work in so many areas, whether it’s financial literacy, climate change like sustainability. I just think there’s so many different sort of big behavioral problems where this sort of combined toolkit could do a lot of good.

Aarthi: I think my biggest aspiration is that funding becomes more flexible to actually allow for it. I think we’ve been lucky to get a lot of sort of flexible ways of putting these projects together, but it’s definitely not the norm. And often folks are trying to piece it together with sort of different components done by different firms. But I’d sort of love to see sort of funding mechanisms come together so that this can sort of be sustained in the field and try it across disciplines and sectors.

Sandi: I agree with everything that Aarthi said, and I think there’s now wider recognition that human behavior is at the core of many of the challenges that we face in many domains of our life. Whether it be helping to improve people to save more or health or even management practices within a business. And that sort of using these principles could improve many aspects of our daily lives. And that would be a really amazing thing.

Brooke: Are there any opportunities that you see out there in the ecosystem to really make a big jump forward in this respect? Do you think that there are specific funding mechanisms that lend themselves well to this? Do you think that there are some conversations ongoing about how to create more dedicated channels for this kind of work to take place? Where are the opportunities out there?

Sandi: Well, one exciting thing that’s happening is that I believe many people are no longer viewing these approaches and tools as being in opposition. So just like we’ve landed on this particular strategy, others have written about variations on this approach. And so it’s worthwhile acknowledging that a lot of other people are arriving at the same conclusion that these tools are actually quite complimentary and could lead to potentially more scalable and sustainable strategies. That’s a really great thing that’s happening.

Sandi: One of the early funders that we had for the project that used priming and social norms was the Bill & Melinda Gates Foundation and specifically their grand challenge explorations mechanism, which has been incredibly flexible and exciting opportunity for many people using these approaches that are thought to be a little bit more innovative, a little bit more on the horizon of behavioral science. And so that’s definitely been one strategy. Another sort of community or practice that I think is important is a group called Design Health that has been organizing conferences and has a website. And they’re trying to be sort of a knowledge hub for all of the activities related to using design in the health domain. But as Aarthi says, clearly these are tools that could expand well beyond health in many aspects of our lives.

Brooke: All right. So on behalf of The Decision Lab and all our listeners, thanks very much for sharing your insights with us today, Sandi and Aarthi. I think for me, my main takeaway from today’s discussion is that we need to pull together the right team to do this kind of interdisciplinary work.

Brooke: Some of the descriptors are the characteristics you mentioned of the right team is that people come with this kind of learner’s mindset or this beginner’s mindset. Starting out from a place of curiosity and humility, really focusing on the problem rather than being married to a specific solution from the outset.

Brooke: And once you’ve got that team together, sticking together for the long haul and giving the team an opportunity to gel and really allowing those bonds of trust to form. So that even when you, as one individual member of the team might not be what it is that one of your colleagues is getting at or what they’re driving towards, you have the faith in them to just sort of stick it out and wait for that thing to become apparent, even if you don’t see it right away. So before we sign off for this week, is there one last thing that you’d like to share either one summative insights for a shameless plug, if you happen to have one, we’re happy with anything.

Sandi: Well, I would say that we’re really curious to hear feedback about the approach that we’ve talked about here today. So we would really welcome folks to reach out to either our LinkedIn accounts or to send us an email, to tell us a little bit about what you think, and maybe if you’re using these same approaches, we’d love to hear about them.

Aarthi: Yeah. I definitely echo that. I think there’s so much potential here. So if there’s folks out there who want to think about using these tools in different ways, or even curious in learning more just love to get in touch and hear from them.

Brooke: That’s great. And for all you listeners out there just want to let that the contact details for Sandi and Aarthi will be on The Decision Lab webpage for this podcast episode.

Brooke: I also want to thank all of you for joining us today. If you’d like to learn more about applied behavioral insights, you can find plenty of materials on our website, thedecisionlab.com. There you’ll also be able to find our newsletter, which features the latest and greatest developments fields, including podcasts, such as this one, as well as great public content about biases, interventions, and our project. Thanks very much for listening and we hope to see you again soon.

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

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