How Integrated Customer Journeys Improve Patient Health

Kailey Raymond: With emerging technologies and multi-billion dollar acquisitions, the landscape of healthcare is changing rapidly. The use of telehealth skyrocketed due to COVID. And tech giant Amazon is set to acquire One Medical for nearly $4 billion. While many see these rapid changes as an opportunity to improve the customer experience, patient privacy remains a concern. Luckily, there are empathetic and data driven marketers like Rahul Kak, VP of Customer Health and Experience at Aetna, a CVS company, at the helm of this revolution.

Rahul and his team work diligently to guide patients towards healthier outcomes through personalized educational nudges, all while assuring their data stays secure and privacy remains intact. In today's episode, Rahul and I discuss privacy versus personalization, how to measure patient outcomes, and how AI models can predict customer behavior. 

Rahul, thanks so much for being here today. I really appreciate it. Would love to learn a little bit more about how you got to where you are today at CVS and Aetna.

Rahul Kak: Well, my life has always revolved around healthcare. My parents used to work in international public health throughout their entire careers. And as a result, I spent a lot of time growing up abroad, mostly India, Indonesia. So in college I studied biomedical engineering, and then interned at the NIH, Medtronic's cardiac devices unit. But I didn't start working in healthcare systems until my second job out of college, which was at CVS in the pharmacy services segment. And then I left for a number of years in between for grad school, consulting at McKinsey, but eventually, landed at Aetna in my current role, which very shortly after got acquired by CVS Health. So my resume comes full circle, but it's been a fun ride to be back in that.

Kailey Raymond: That's interesting. So you were inspired in a lot of ways by your parents, but you also are like a boomerang employee to CVS specifically, with a stint in consulting. I'm curious to learn a little bit more about some of those jumps and some of those learnings. What have you learned with consulting that, now, you're bringing into healthcare, and what are some of those paths that you're weaving together?

Rahul Kak: Well, I think the skill sets I got from being an engineer by my training and then also as a consultant was problem solving, and I just think that is so important to be a leader in any industry. You want to be a really whip smart problem solver and be able to figure out a framework to put out fires. So I think that's key. There's a big part of my previous life as a consultant, just having a network. And even just from graduate school, just having a support network to feel like you can ask dumb questions to, because there's a limit for how much you can ask that when you are new to an industry or a sub-sector of an industry. Just having a safe space to be able to learn.

Kailey Raymond: I like that. I like that answer, as consultants are always on the job. Their entire job in a lot of ways is to learn. And so coming in and being able to have that ability to cast out questions to a network of people and play the dumb one in a lot of ways to gain as much information as you can. Curious specifically about your role at CVS Health right now, what is your team's role? What are you in charge of? What are some of the activities that you think about all the time?

Rahul Kak: So my title is VP of Customer Health and Experience. And I sit within the marketing organization of our health benefits business unit, but probably 85% of my work is in health benefits and about 15% is on pharmacy services. So what I do is I lead an agile marketing department of cross functional marketers that are matrixed in with analytics, data and clinical colleagues, and we are developing these highly targeted communication nudges that are all about empowering our customers with personalized information to help them make better health decisions.

And so we're doing this for our Aetna medical members and our Caremark pharmacy benefit members. I would say it's not marketing so much as it is population health management, but it is using marketing channels. And this can happen at all points in the lifespan of disease management, from prevention to early detection, downstream to reducing complications. Then we also have non condition specific stuff, for example, helping our members navigate to higher quality, low cost sites of care.

Kailey Raymond: Wow. High stakes industry, absolutely. And as I understand it, y'all cover 40 million members or more. Is that right?

Rahul Kak: It just depends how you cut the data. The 40 million would be if you include Aetna Medical plus pharmacy benefits. Talk about CVS Health Enterprise, it's over 135 million unique customers in the United States, across all of our business units. A third of the US alone is covered by our prescription benefits division. But then the medical benefits are about on the order of 20 plus million Americans. So whatever it is, it's big numbers.

Kailey Raymond: Wow. That's amazing. And I'm so excited to dive into some of the specifics around the campaigns. It sounds like there's a lot of life cycle management for the marketers in the room that might be... the terminology that we'll throw out there. You're in the business of health. Stakes are obviously high, and with that scale, I'm sure that you're looking really closely at some of the trends within the industry to make sure that y'all are on top of it. So what are some of the top trends that you're seeing as it currently relates to marketing and customer engagement in the healthcare industry specifically?

Rahul Kak: Sure. I'll give you three. Consumerization, vertical integration and price transparency. And I'll try to talk about how that impacts consumer engagement. So number one, the healthcare payer industry has traditionally been very B2B focused. You would win business by winning big employer groups, but as you see Medicare Advantage exploding, individual exchanges like healthcare.gov exploding, Medicaid, you're seeing a shift with healthcare becoming increasingly consumerized in every one of us and our competitors, we want to be consumer businesses and you're seeing-

Kailey Raymond: B2B2C.

Rahul Kak: Yeah, exactly. And there are elements of the business that are straight B2C, and you are even seeing providers experiment with direct to consumer models. So that's important because marketers are at the heart of that. We often like to call ourselves bringing the voice of the consumer to the business that we sit in. That's number one. I think our role as marketers are increasingly important in healthcare. Number two, vertical integration. I mention how CVS Health bought Aetna a couple years ago, back when I worked for Caremark. I joined two years after CVS bought that. And this is happening all across the industries, big guys are buying other elements of the system. Because what we are all trying to do is create a better, more integrated experience in the patient journey, have fewer handoffs. And on the marketing side, you're going to see that play out with more investment in technologies, that weave our customer communication platforms onto a single tech stack, CDPs that are bringing all that data together and allowing more coordinated, smart orchestration across all of our communications. That's trend number two. And as a marketer, you want to be on top of that, you want to understand the martech stack, you want to understand the players.

Kailey Raymond: Absolutely.

Rahul Kak: And if not how to use the tool, how to work with people who know how to use the tools.

Kailey Raymond: That's key. I actually am curious to dig into the second one specifically, which is obviously, maybe a little bit self-interested perhaps, but what I'm really curious about is, obviously this big acquisition happened, what, 2018 CVS acquires Aetna. And I'm sure that you are, for many years, maybe still, integrating data systems. Interoperability between all of these different systems is going to be crucial. What does that process look like, especially from making sure that you are bringing together all of the right data, the same patients, making sure that you understand everything about them and their journey? What has that been like for y'all?

Rahul Kak: I think it's, until you have all of your systems talking, it's a lot of manual coordination. And it takes years to go through a proper merger for companies of our size, where all of your systems are fully adapted to each other. Especially in healthcare, we have a lot of legacy systems, legacy infrastructure, where just as an extreme example, for listeners who are in big companies, you may have only one person who knows how to operate this one single campaign, because they're the only ones who understand the data flow from a product team. And it's just crazy stories like that, that you hear, where there isn't fungibility between people. I think it's important to identify where are those gaps and work towards building a cohesive infrastructure that can allow a data fabric layer where things can pass through, and you don't have these single points of failures where you're relying on individuals to run certain programs. So that's why it's important. And some of the challenges we've faced over the years, broadly as a company, to bring our systems together and bring a better patient and member experience out there. 

When you talk about multi-channel coordination for automatically triggered nudges, there are systems in place that can orchestrate the messaging to make sure people are not over messaged or sent conflicting information that'll happen through data. And that's super important when you have a vertical integration where you have maybe people touching different points of a patient journey. There's a high risk of, especially with hundreds of teams that are trying to support similar objectives, that they can step over each other's toes. And we have also had multiple governance functions where our teams are reviewing communications to make sure our messages are strategically aligned, on brand, coordinated, catch duplication or redundancies before they happen. So always something we're trying to get better at as a big company.

Kailey Raymond: What I heard, what I want to dig into a little bit deeper, is different channels. Making sure that we can touch the right customer at the right place at the right time is something that we definitely love to learn more about, and making sure that you are using channel of choice and that you're using contextual data to make sure that you're speaking to the customer in a way that they really trust and believe and want to act.

So tell me a little bit more about what you're thinking about at CVS Health, as it relates to leveraging these different channels that you have at your disposal. First of all, what are they? And second of all, maybe walk me through what a customer experience might be like across a couple of different channels, maybe text and pharmacy or something.

Rahul Kak: First off, I'll say in our industry, probably the best channel that you would have is an actual person, especially for such a complicated space as this. Ideally, you have a care manager or some sort of clinician that is jumping on the phone with you. And this is common at a lot of health plans or managed care plans that actually hire staff of clinicians and nurses, or even just case managers that just really know the space, that can be coaches for people. And that will always be the most successful form of behavior change in health management. 

However, it is not super scalable, because there's a limit to how many people that you can hit. So, that full level of scalability, you're going to get through marketing channels. That's getting to what we were talking about, things like direct mail, email, text, in addition to live phone, automated, outbound phone calls, mobile app notifications. We will create micro sites that are personalized one to one.

Kailey Raymond: For individuals, you'll have a micro site for me?

Rahul Kak: Correct.

Kailey Raymond: What does that look like?

Rahul Kak: That would look like, you click through, we would have the data to know what address we have on file for you, and therefore, as an example, "Here's some sites of care that we would recommend that are high performing, convenient providers that are in-network that are in your area." And then you can also get into more specific things. It may be a customized message based on condition, but you have to be careful because there's a very interesting world of towing the line with PHI, personal health information, and understanding what's a secure channel and what's an insecure channel. So the one-on-one personalized micro sites I described, you can't get too specific unless you create an authentication layer that they log in, or at the very least say, "Here's my date of birth. Here's my zip code."

Kailey Raymond: Interesting.

Rahul Kak: If that makes sense.

Kailey Raymond: Yes.

Rahul Kak: And a lot of people don't realize, the government, the law, would say that emails and text messages are not secure channels, meaning we can't actually be very specific or very personalized there in the way we want to be in our industry. So you need to go to enveloped mail, authenticated phone calls, and I'd mentioned authentication layers, things like that. You have to be really creative as marketers to manage that.

Kailey Raymond: No kidding. And so, I guess what I'm learning a little bit, is you need to be logged in for you, CVS Health, to be able to provide to your ability, the best patient outcomes or the best experience. Is that right?

Rahul Kak: Yeah, that's right. And then there's other channels, I forgot to mention. We have an Apple Watch program, and that's again, that's a digitally... and it's pre-authenticated when you're using it. In some cases you can go to a CVS store, you'd have a pharmacist who can be pinged for a nudge that's specific to you as an Aetna member. So you can get a sense for how we're integrating our businesses. But otherwise yes, broadly that's right. We do have to be really thoughtful about if you're talking to someone who is a diabetic or someone that you think is at risk for having diabetes, you'd have predictive models that indicate it, but you can't specifically say the condition because it's not secure. Or at least the law wouldn't say it's not secure.

Kailey Raymond: Interesting. So what does that messaging look like? If you think you know something about somebody, do you wait until perhaps they visit the pharmacy and then have that real time context up in front of a pharmacist to deliver that message in real life? Or how does that look for you?

Rahul Kak: You see pharmacies that actually have a bit more flexibility. And you've probably gotten this before, you get a text message from your pharmacy that says, "Your drug," and it's maybe three letters of the drug and you have to figure out what that means. You understand why that happens now. So what you want to do as a marketer, is you want to weave together the channels so you can at least have the initial communication, help them understand maybe there's an important message, or maybe general health education and say to learn more specific about you, you might have to ask them specifically to log into their app or on the website or provide some sort of authentication information that allow us to deliver the secure channel. Or you spend a lot of money on print, because envelope mail is considered secure. So you have to think about your whole palette of colors that you can use.

Kailey Raymond: Wow. That is so interesting. And I just can't get past this point of you being able to provide the best outcomes for folks if they have a logged in experience, I think is such an interesting takeaway for folks. Just to manage your health in so many ways and make sure that you are getting the most out of your team's work, and the predictive analytics that are going into place, being in the secure channel is important, right?

Rahul Kak: Yeah. It is important. I'll brag for a sec. That'll make you feel happy, where I did listen to some of your previous podcast episodes, and there's one individual, I think it was the NBC Universal guy who talked about the paradox between privacy and personalization. And when you talk about health data, that is one of the most valued pieces of data you have about yourself, and there is a lot of concern about privacy there. We're going to be very thoughtful about that as stewards of that data, of our knowledge about your medical or your prescription behaviors. But we want to use that data to personalize messages to help you manage your health. So we do have to be really conscious and work very closely with our regulatory partners, but also just think as consumers, about what's the best way that we can help our folks by being personalized, but being stewards of their private data. Kailey Raymond: That's interesting. Privacy creates this very obvious tension in so many ways in every industry. And there's so many layers to it, but specifically in healthcare, it seems like it's even more nuanced. I guess I want to circle back, because you've mentioned now a couple times as the first trend that you mentioned, is consumerism. And a lot of companies, either going DTC or figuring out B2B2C models. Are there any consumer behaviors that you've seen in the last couple of years that might be really driving that trend from a B2B perspective? What have you seen within healthcare that's really making sure that people are going out actively and transitioning models from B2B to B2C?

Rahul Kak: So transitioning from B2B to B2C, I'll give you two trends. One is around what we've seen around usage of digital tools. Like our digital tools are being used far more on average over the past couple years, whether that's used for a doctor, estimating costs, reviewing coverage. All of these different sort of transactional activities are increasing all the time. And you can imagine a lot of that was happening because of COVID. COVID's been a huge catalyst for all sorts of change, and accelerators for all sorts of things in the healthcare system as a whole. But digitization is a big part of that. Another related piece and related to COVID, is people have gotten much more comfort in utilization of telemedicine services, particularly spiking heavily for behavioral health issues, whether you're a payer, provider, you're seeing this spiking in utilization of digital services.

Your ability to communicate with your members, your patients, is a lot more heightened when you have increasing use of those. So I think being really thoughtful about what's that full experience. And again, making sure it's coordinated between all of your channels internally, and increasingly more of a desire to see coordination across different players in the health system. So I think that's one of the reasons you actually see vertical integration where someone who's a big player in one space, they want to own other parts of the health experience journey, because they know if they own the pieces on the board, they're able to connect them more seamlessly.

Kailey Raymond: All of these things are piecing very nicely together, right? Consumers are driving this trend towards digitization. That is certainly the leading companies towards this trend of really thinking about what their martech stack looks like and making sure that they are creating a composable enterprise to be able to bring in and out tools and replay that data to make sure you're never losing anything in the customer experience and journey.

I'm curious, and this is kind of an existential question, so feel free to take it where you want to. But if you could, do you have a definition for what good data looks like for CVS Health?

Rahul Kak: What good data looks like. I would say good data is structured and it's consistently captured and consistently manipulated in such a way that you're able to ingest it in different systems. Especially a very large company needs to be able to pull it into different models, either for targeting, personalization, measurement. If you want to connect it with other vendors and systems, and there's so many innovators in the healthcare space, but also in the marketing space, you need to have structured and just clean data to be able to actually connect the pipes. Otherwise, you're going to have a lot of integration headaches to create simple adapters or APIs.

Kailey Raymond: It's really the foundation in so many ways, in this new world of digital experience to make sure that you are speaking the right language to people at the right time in a way that they want to be spoken to. You mentioned one very personalized tactic, which I didn't realize you were doing, which is one to one landing pages, which is so cool. What are some other examples of personalization in the healthcare space that you all are either working on or thinking about?

Rahul Kak: I would say personalization could happen at a few levels. You have the top of the funnel, which is, think about use it for targeting, which I mentioned. So we could use, for example, machine learning to suggest there's a higher probability this message is relevant to you, so then we'll send you that message. But if you're getting that message, you can still personalize that to different groups. So, go a level down in the funnel, it's a little bit tighter. I'll give you an example. Say you're trying to reach Medicare women and you know a large percentage of them are not getting mammograms, you can put them into different cohorts. One cohort is someone has too many chronic diseases to even think about getting their screening. They're just managing too much and their brain is overloading. And you have another cohort that's not getting their mammogram because they're just physically too far away from a screening clinic.

And you can imagine, as a marketer, you're creating very different messaging for those two different types of cohorts. Then you go to the bottom of the funnel, and yes, you can personalize messaging one to one, based on information you have. And I gave you the example that was based on addresses and then based on conditions. Other things you could have, you could be based on language. There's a big talk about health equities and how can you actually create culturally concordant messaging. Again, that's not one to one, but again, think of different cultural groups and people who speak different languages as their own cohort or a cohort that layers on top of these other more medical related cohorts I talked about, or behavior based cohorts.

So you want to think about all of those things. And then you just get the simple things where it's like, say, someone's name. Then if you do an A/B test, oftentimes you find just saying someone's name, if you have that information, give you a tiny little incremental lift on open rates. Maybe not behavior change rates on health, but again, you want to have different points of measurement across their journey, from engagement through actual conversion or behavior change.

Kailey Raymond: What are those things specifically that you are measuring? Because at the end of the day, what we're really talking about is people being healthy, and marketing is a massive influence to how they're getting there, clearly. And so, open rates don't make me feel warm and fuzzy inside when the outcome is like, "Yeah, I'm a healthier human being." So what are some of the ways that y'all are measuring some of those patient outcomes in the end?

Rahul Kak: Yeah. And on the payer side, it's interesting. We get medical claims in and pharmacy claims, and we have a lot of data about claims from providers that are asking for reimbursement. So we have a lot of knowledge about the cost, the total cost, of the managing the system. So in doing that, we can look into what is the most efficient sites of care in which we are trying to reduce the amount of out of pocket you can spend as a consumer, but then it's also the counts of how many people are getting colon cancer screenings, mammograms, vaccines. There's a famous set of standard measures that's called HEDIS, which stands for healthcare effectiveness data and information set. And that just provides a whole bunch of standardized things that's used across payers, across providers, across the industry, that's used to measure health outcomes and quality metrics. And if you don't know, providers and payers are both often reimbursed by the government with bonuses for actually hitting those measures. There's a lot of incentives in play to get us to do the right things. 

Other types of things we would think about would actually be perceptions of access to care. And this is just survey based information. You're just reaching out to people and there's a healthy outcome survey. There's another one called CAPS. It's about a consumer assessment of how much they think they are able to access their health. And that's important. People will tell you, "Do I feel like I have access to a PCP or a primary care provider? How long does it take to get them?" And that wait time to reach someone means that you can be a major barrier for someone getting them. So those are some of the examples that we'll think about for impact measurement.

Kailey Raymond: That's really interesting. And one of the things that we've spoken about a couple of times on this show too, is exactly like bridging that gap between survey data, getting direct from a human being on the other side, them telling you what they actually believe or seeing and feeling. Click data, maybe they're logged into their personalized landing page and they're working around and figuring out what's on their apps. How are you bringing together those two different signals for one person to be able to say, "Kailey believes this and needs this. And this is the way that we're going to further her on her care journey"?

Rahul Kak: We're using less of the survey data than we are necessarily with the hard data, things that are more like actual behaviors and deterministic things. But there are data points we can get with perception. And we'll do this sometimes where we can augment some of our targeting models and our personalization models, because there's data lacking in the system, by just reaching out to people. We actually did this with COVID vaccinations, where we didn't actually have information, whether or not people were vaccinated from the government, it's just been sort of messy to know for sure is this person in our membership actually vaccinated?

Kailey Raymond: I'm glad to hear you say that. It's felt a little bit messy on my end, but I'm glad to hear that it is. Okay.

Rahul Kak: I mean, all these things had to happen so quickly. We had that experience with COVID, where we were actually reaching out to our members in certain lines of business. And we were just trying to do a sample size across a certain sample size, not the whole population, trying to find out are they vaccinated or not, and if not, why not? And then we were using that to actually augment our missing data. But then we're actually using that to feed into our models and figure out like, "Okay, based on that, can we create a lookalike model to find other people who are like this group of people who said, 'These are the reasons I was not getting vaccinated'? And you use those lookalike models to develop different marketing treatments or really, just sort of behavioral nudges, to try to get people to think about things differently. So it's that kind of two effects. One, augmenting missing data. And then, two, building in AI driven models.

Kailey Raymond: That's so interesting. Do you ever use those AI driven models, in the case of COVID, maybe seems a little bit less scary. I've certainly seen some of those big campaigns. Do you ever use any of that predictive data and cohorts of folks for certain educational campaigns?

Rahul Kak: Yeah. Educational campaigns. But I can for sure say that we are using AI to build targeting models that might predict future behaviors or future risks. So for example, there is a probability someone is at risk for substance abuse. There may be a high likelihood someone may be getting a certain procedure, and what we know about those procedures is that there are high quality, low cost options in your area. We could be using it to identify a higher probability that someone might be undiagnosed with an unusual condition that won't be caught unless they go to a specific specialist, and then you can nudge them to go like, "Add this type of specialist to your care team." But we have to be really careful about these types of AI driven nudges, because we'll never get 100% precision on those models, but it creates an incredibly interesting world for marketers to play in. And when I say marketers, I mean, people who are managing the customer, member, patient engagement journey.

Kailey Raymond: That's so interesting, as in so many ways. It's like AI saved my life, or it could be, where you're in this cohort for a certain reason, you're building predictive models on your teams and then you get an email that you decide to open one day that educates you into something that you'd never even really considered before. That is inspiring. That's a really amazing line of business and way to use data. 

What I'm wondering is, do you think that anybody's doing it right in terms of customer engagement? I'd actually love if you have any examples of folks in the healthcare space that you think are doing things in a really great way.

Rahul Kak: I don't have many examples in the healthcare space, and I should probably pay more attention to that. It's funny, as a consumer, it's so easy to gravitate who's at the top of their game. I'll take another example that I heard in one of your previous episodes. My favorite consumer campaign is the Spotify Wrapped campaign.

Kailey Raymond: You're stealing my stuff, Rahul, you're stealing my stuff.

Rahul Kak: I'm stealing you stuff, but I'll tie it back to how I think about. One, it's like they're using behavioral data to culminate in an interesting presentation of data that's specific to you, so it's like one to one, which is something we're always thinking about in healthcare. But then they also take the funniest, interesting stories and they put it up on billboards around New York and presumably other types of mass media. So what they've also done is they've created a multi-channel campaign that's not only going viral on digital channels, but it's also synced to out of home, and that's truly impressive. And that's something I think about a lot as a marketer, is how can we utilize all of our channels to create that experience that feels surround sound?

Kailey Raymond: I hear you. Spotify Wrapped is truly the best part of December in so many ways for me, to be able to look back on my listening data and probably be pretty embarrassed by what I'm seeing, frankly. It's like they know me better than I know myself in so many ways. And I don't know if I would enjoy that from a health app, I will say that. So I do think that there's... What I'm getting is there's just so many nuances and there's so much trust and there's so much privacy, of course, related to this field specifically. But the ability to make change in somebody's life is so powerful. It's just a really amazing field to be in. And there's a lot of really creative solutions that I'm sure y'all have to think about every single day as it relates to messaging, to get the impact that you're hoping and desiring. What I'm wondering is, do you have a favorite piece of data or a favorite data driven marketing campaign?

Rahul Kak: Let me give you one on hospital readmissions, and traditionally what many health plans do, managed care plans, is that they will call you after a hospital discharge to make sure you're taking care of yourself and reducing the likelihood of getting readmitted. There are actually penalties from the government for providers that have a high readmission rate, which there should be. It means you probably didn't do your procedure right, or something like that. There's a lot that can go wrong. What we found is that marketing communications can be almost as effective in some of these traditional payer outbound call programs. As I'd mentioned before, it is way more scalable to use marketing channels than live humans. What you want to do is augment them.

And in one of our campaigns about that, which is just educating people about what to do, either prior to your hospital admission, if we know ahead of time and if it's scheduled, and what to do afterwards. And we were seeing as high as a 15% behavior change rate and reduction of readmission rates across certain windows. It's crazy to think that we're using well-timed mailers and digital nudges to keep people out of the hospital, and the numbers are so good on that. And the answer is, "Yes, it works," because what we found afterwards, what we were doing is we were actually just nudging people to go see their PCP and know what to talk to their doctor about post discharge, and that was highly correlated with the reduction in hospital readmission. That's not a secret correlation by any means, but the fact that we were able to use simple marketing channels effectively get to that outcome is incredible.

Kailey Raymond: That's amazing. And I mean, I think that's... So many people, myself included, I don't always feel equipped to know exactly what questions to ask my doctor. So, you being the experts, being able to educate people at the right time is amazing to be able to build that trust too. And that's something that we talk about all the time here, is there's this big perception gap between what companies think in terms of their customers trusting them and what actually is going on. And that's obviously huge in this industry, but to be able to point to something like that, that feels like you're on my side. That feels like you are knowledgeable and trusted in what you do, and so that is a huge way for you to be able to have more people join your app, et cetera, in the future, and get all of that data to drive better outcomes. What changes do you see on the horizon for customer data and experience in the next six to 12, or more months, as it relates to healthcare?

Rahul Kak: I think you will continue to see that look back trend I mentioned before that will continue a vertical integration. And I think you're going to see payers integrating more with primary care practices, and a lot of it through telemedicine, which is super hot right now, especially the concept of virtual primary care. I think it is the next big wave of vertical integration in our industry. Whereas you look back 15 years ago, probably the big trend then was health plans merging with pharmacy benefit managers. CVS, ourselves, we've announced our intention to extend from our MinuteClinic and HealthHUB platform into primary care. You've seen Amazon start to catalyze it with their recent acquisition of One Medical. And I think this is an exciting change in the industry to see a heavier focus on primary care because primary care is all about prevention, and it's coaching people into healthier lifestyles. Despite primary care physicians being the lowest paid in their field, it plays such an important role.

And there are certainly, everyone would agree, not enough, family doctors out there in the country. So the more that we can do in the healthcare space to make primary care accessible, the better long term consequences we'll have. And as marketers, we're going to have to sell that story and get people to utilize those services when it becomes part of the benefit package. That's also going to be a big push, even when you talk about virtual primary care. That is so powerful, that is so convenient, but it's also unnatural, the idea that my doctor's on a screen. And it's natural for some specialties, I'd mention behavioral health, it's heavy utilization on that. But if we can improve the utilization of those types of packages and primary care services, as they become more available, then we as marketers are doing our part.

Kailey Raymond: Amazing. Last question. What are the steps or recommendations that you might have for somebody that's looking to up level their customer engagement strategy?

Rahul Kak: I would say ask questions and don't take things at face value. You should always be asking about why do we do this thing this way? Especially when you work in a very entrenched industry like healthcare, or just a very large, big company that you have a lot of entrenched players within that. And I think the first step to building great experiences for your customers means your internal operations and how you get work done has to be efficient and it has to enable you to innovate. So that's number one, ask questions, don't take things at face value. Number two is experiment, and this is just creating a culture of test and learn. And that is simply synonymous with modern marketing. If you have the expertise, use randomized control trials, when possible. Don't limit yourself to A/B tests. Can you do a multivariate test? How many tests can you feasibly cram in where you can still measure a statistically significant outcome? Just test, test, test.

Kailey Raymond: I love it. This is amazing. Thank you so much for your insights, Rahul. And go see your primary care doctor, I think, is probably like, "I got to make that appointment now." So always a good reminder. Appreciate you being here.

Rahul Kak: Thank you.