How WHOOP Uses Big Data to Optimize Your Fitness and Health

Most fitness gadgets, like the Fitbit or the Apple Watch, encourage you to get out there every day and “close your rings” or “do your 10,000 steps.” But there’s one activity tracker that’s a little different. The WHOOP isn’t designed to tell you when to work out—it’s designed to tell you when to stop. 

Harry’s guest this week is Emily Capodilupo, the senior vice president of data science and research at Boston-based WHOOP, which is based here in Boston. To explain why the company focuses on measuring what it calls strain, rather than counting steps or calories, she reaches all the way back to the beginning of the company in 2012. That’s when founder and CEO Will Ahmed had just finished college at Harvard and was looking back at his experiences on the varsity squash team. Ahmed realized that had often underperformed because he had overtrained, neglecting to give his body time to recover between workouts or between matches. To this day, WHOOP designs the WHOOP band and its accompanying smartphone software around measuring the physical quantities that best predict athletic performance, and giving users feedback that can help them decide how much to push or not push on a given day.

Capodilupo calls the WHOOP band “the first wearable that tells you to do less.” But it’s really all about designing a safe and effective training program and helping users make smarter decisions. Meanwhile, the WHOOP band collects so many different forms of data that it can also help to detect conditions like atrial fibrillation, or even predict whether you’re about to be diagnosed with Covid-19. It’s not a medical device, but Capodilupo acknowledges that the line between wellness and diagnostics is shifting all the time.  And with the rise of telemedicine, which is spreading even faster thanks to the pandemic, she predicts that more patients and more doctors will want access to the kinds of health data that the WHOOP band and other trackers collect 24/7.

The conversation touched on a very different way of thinking about fitness and health, and on the relationship between big data and quality of life—which is, after all, the main theme of the show.

Please rate and review The Harry Glorikian Show on Apple Podcasts! Here’s how to do that from an iPhone, iPad, or iPod touch:

1. Open the Podcasts app on your iPhone, iPad, or Mac.

2. Navigate to The Harry Glorikian Show podcast. You can find it by searching for it or selecting it from your library. Just note that you’ll have to go to the series page which shows all the episodes, not just the page for a single episode.

3. Scroll down to find the subhead titled “Ratings & Reviews.”

4. Under one of the highlighted reviews, select “Write a Review.”

5. Next, select a star rating at the top — you have the option of choosing between one and five stars.

6. Using the text box at the top, write a title for your review. Then, in the lower text box, write your review. Your review can be up to 300 words long.

7. Once you’ve finished, select “Send” or “Save” in the top-right corner.

8. If you’ve never left a podcast review before, enter a nickname. Your nickname will be displayed next to any reviews you leave from here on out.

9. After selecting a nickname, tap OK. Your review may not be immediately visible.

That’s it! Thanks so much.

Transcript

Harry Glorikian: Hello. I’m Harry Glorikian, and this is The Harry Glorikian Show, where we explore how technology is changing everything we know about healthcare.

If you’re a gadget lover and data aficionado like me, you’ve probably tried a lot of different fitness monitors and other wearable devices, like a Fitbit, or an Oura ring, or an Apple Watch.

We’ve talked about a lot of these devices on the show. Usually they come with a smartphone app, or they run their own apps.

And the job of the apps is to track your fitness progress and encourage you to get out there every day and “close your rings” or “do your 10,000 steps.”

But there’s one activity tracker that’s a little different. It’s the WHOOP band.

The WHOOP is not designed to tell you when to work out. It’s designed to tell you when to stop.

My guest today is Emily Capodilupo. She’s the senior vice president of data science and research at WHOOP, which is based here in Boston.

And to explain why the company focuses on measuring what it calls strain, rather than counting steps or calories, she reaches all the way back to the beginning of the company in 2012.

That’s when founder and CEO Will Ahmed had just finished college at Harvard and was looking back at his experiences on the varsity squash team.

I’ll let Emily tell the whole story, but basically Will realized that had often underperformed because he had overtrained, neglecting to give his body time to recover between workouts or between matches.

To this day, WHOOP designs its signature WHOOP band and its accompanying smartphone software around measuring the physical quantities that best predict athletic performance, and giving users feedback that can help them decide how much to push or not push on a given day.

Emily calls the WHOOP band “the first wearable that tells you to do less.”

But it’s really all about designing a safe and effective training program and helping users make smarter decisions.

Meanwhile, the WHOOP band collects so many different forms of data that it can also help to detect conditions like atrial fibrillation, or even predict whether you’re about to be diagnosed with Covid-19.

But it’s not a medical device.

But Emily acknowledges that the line between wellness and diagnostics is shifting all the time.

And with the rise of telemedicine, which is spreading even faster thanks to the pandemic, she predicts that more patients and more doctors will want access to the kinds of health data that the WHOOP band and other trackers collect 24/7.

It was a fascinating conversation that touched on a very different way of thinking about fitness and health, and on the relationship between big data and quality of life, which is, after all, the main theme of this show.

So I want to play the whole interview for you now.

Harry Glorikian: Emily, welcome to the show.

Emily Capodilupo: Thanks so much for having me.

Harry Glorikian: Yeah, I have to tell you, I was reading your background and I’m like, oh, my God, I’m so excited. She comes from like, you know, like real training in sleep. And we’re going to talk about these devices. And it’s one of the things I use them all for, as you can tell, like I’m I’m sort of geared up and I’ve got all of them and I and I cross correlate and I can tell when somebody has updated something and the algorithm, like I can see like all of a sudden they start moving apart from each other or being different from each other. But, you know, for those people who aren’t, say, up to speed on the world of fitness monitors, I’d love for you to start, you know, by explaining you WHOOP’s mission, and then maybe talk about different parts of your system, you know, like the band, the sensors, you know, the basic capabilities, that sort of stuff.

Emily Capodilupo: Sure. So WHOOP’s mission is to unlock human performance. And in a lot of ways it started out at the beginning. You really focus on athletic performance. Our origin story is very much in preventing overtraining. But as we started to do more and more research, we started to discover that the things that predict athletic performance at the sort of root physiological level are actually the same things that predict all kinds of performance. So we’ve seen them predict things like cognitive performance. We’ve seen them predict like emotional intelligence and, you know, like how short you are with people, stuff like that, you know, as well as like how people feel like they’re performing at work or in their jobs, in their relationship, stuff like that. So while …physical performance is, where a lot of those algorithms and sort of like our research started, we started to realize that without tweaking any of the algorithms at all, they started to be really good predictors of other elements of performance as well. So we’ve really broadened our mission. It’s all about unlocking human performance in the broadest sense possible, and we do that with this device. Some of the things that we think are really important about our design as it compares to some of the other wearables, is that as you’ll see, it’s screenless. And we really think about the device just as this itty bitty little bit that slides out from the fabric.

Emily Capodilupo: And so it’s actually capable of being worn almost anywhere on your body. So we have clothing that totally hides it. You can wear it in your underwear, on your bra, on a t shirt, anything like that, as well as sort of the traditional wearable locations like on your wrist or bicep. And one of the reasons why we wanted that form factor is we really wanted to collect 24/7 data and be able to get this complete picture of your body. It actually charges wirelessly so you don’t even have to take it off to charge it. And that allows us to get the most complete picture of what’s going on. And so we don’t miss like the 2 hours when you take it off to charge or you don’t charge it overnight and then miss the sleep or anything like that. So it gives us this like really incredible picture. Kind of one of the other important differentiators just in the hardware itself is because we’re not powering a screen, we’re able to put 100% of the battery into driving the sensors and getting the most accurate signal. And so when you start with the most accurate signal, the most accurate raw data, you’re then able to power better feedback, better coaching, because you’re starting with something more reliable. And so we’ve done a lot on the coaching side and the algorithms side that other wearables just haven’t been able to do.

Harry Glorikian: Interesting. So Will Ahmed and John…and I’m going to try to pronounce it.

Emily Capodilupo: Capodilupo.

Harry Glorikian: Thank you. Started WHOOP in 2012, right? While John was at Harvard and Will had just graduated. Right. So, you know, I mean, maybe a little bit about the company’s origin story or. I don’t. God, that was you know, if I go back that far, the fitness monitoring market was like in its nascency.

Emily Capodilupo: Yeah it was, the Jawbone Up had just come out, the original Fitbits had just come out. And not too long after that the Nike FuelBand started, which no longer exists, of course. And, you know, if you look at what wearables were doing at the time. Oh, and then, of course, there was this other class of wearables that had been around for a little bit, which were like the Garmin running watches. So it kind of GPS watches that you put on for the run or for a bike ride or whatever it is. It would capture all the GPS data, give you information about your pace, and then you take it off when the run was over. And so you kind of had those like two classes of wearables. We had these like 24-ish/7 step counters, and then you had the like more intense while you were working out data, but nobody was really bridging those things. But the sort of theme across all wearables, both of those different categories at the time, was this like push harder, more is more, faster is better, just do it, right. All of those kinds of messaging. And we weren’t really seeing, at least with the like kind of step counter class of wearables, we weren’t seeing any kind of adoption in like elite athletes or even like collegiate athletes because they didn’t really need to be told do more.

Emily Capodilupo: And actually what happened is, sort of the WHOOP origin story is, Will was captain of the Harvard squash team. And when he got named captain, he sort of committed that “I’m the captain. I should work harder than everybody else. That’s what a leader does.” And he worked so, so hard that he overtrained, really burnt himself out and like did really poorly. And he had this moment of like, you know, I’m in a Division I school and I’m like the fanciest, you know, squash programs that there is. How come nobody knew I was overtraining and like, told me to stop. And like, who knew that this was a thing? Like, I always thought that if I worked harder, I’d get better. And actually, you can work too hard and working too hard is bad. And he found that like everybody on his team was really motivated to work hard and sort of motivating each other to work harder. And they didn’t have that balancing voice of like, Oh, I should take a rest day and like sit out, even though like my teammates are practicing. That would have felt like very uncomfortable and like not being a team player or something like that. But he started digging into the data and it really did show that like actually when you need a rest day, you will be stronger for having taken the rest day, than you will be for like manning up and pushing through.

Emily Capodilupo: And so he really set out to create the first wearable that was going to tell you to do less. It was very countercultural in that moment. But he was trying to address kind of the highly motivated market that needed almost like permission to pull back and to be told what their limits were. And so from day one, we were really focused on like, how can we create a recovery score that’s going to tell you, like, you’re better off resting today than you are like doing this program or that, like, a coach could use and see the data and say, okay, these four players, they’re going to do an extra set or an extra drill or whatever it is. And these four players, they’re actually going to stop 20 minutes early and, you know, go sit in the sauna or stretch or whatever it is. And by modulating people’s training in response to their bodies, readiness to respond to that training, actually create like safer and more effective training programs. And that was where we started and then kind of evolved into the product we are right now. But a lot of that is very, very much, that philosophy is still kind of at the core of what we’re doing.

Harry Glorikian: Yeah, I definitely have questions. We definitely have to talk about the recovery score and sleep apnea, because I have a vested interest in understanding this better. Actually, it’s funny, I try to talk about this with my doctor and he’s like, “Man, you know more than I do about this.” But so, you know, thinking about how the company is evolving. It’s been moving forward. I’ve been watching it. I mean, what is the company’s sort of larger philosophy about like the role of technology in fitness and health. I mean, do you feel like we’re headed towards a future where everybody is going to rely on their mobile and wearable devices for health advice?

Emily Capodilupo: I think so. And I think that, you know, there’s a big asterisk to that answer, which is I don’t think that wearables are ever going to replace doctors, and I don’t think that we’re trying to do that either. But we do have a lot of information that doctors don’t have. And there’s a really, I think, exciting opportunity if the medical community were more open to it. And they’re definitely shifting in that direction. And that’s been accelerated by the pandemic and the rise of telemedicine, where there really is an opportunity. I mean, if you think about it, just like the really simple basic stuff like telemedicine appointments skyrocketed during the pandemic.

Harry Glorikian: Right.

Emily Capodilupo: Every other in-person doctor’s appointment I’ve ever been to, the first thing they do is they take your vital signs right, often before you even get to see the doctor. They’ve taken your vital signs, or if you’ve a telemedicine appointment, they just totally skip it, right? And so it’s like, well, you know, my wearable can tell you what my resting heart rate is, could tell you not just what it was this morning, but what it’s been all month and all that kind of stuff. It also can tell you what my blood oxygen level is, my temperature. And that’s a lot of information that’s like, you know, is a lot better than having nothing. Which is what telemedicine has right now. And so it’s not like let’s throw out all the EKG machines and all of that.

Emily Capodilupo: But, you know, there are a lot of situations where remote monitoring can add a lot of value. And then there’s other places where even if the doctor was there to take your vital signs, sometimes vital signs in context have a lot more information than an isolated reading. So like we published a paper about a little over a year ago now where we were looking at respiratory rate in response to COVID-19 infections. And what we found was about three days before or up to three days before reported symptom onset, people’s respiratory rates were starting to climb. And we would see this like because daily your respiratory rate when you’re healthy, it doesn’t change at all from night to night, it’s super flat. And so it will be like the exact same thing night after night. And then all of a sudden you’d see this spike like two, three days before COVID-19 symptom onset. It would stay up or keep climbing. And then three days later, people would say, like, Oh, I don’t feel well, whatever. They go get a COVID test, and lo and behold, it would be positive. And so it was this like interesting early warning sign. But what was really, really interesting about that study is that oftentimes people’s respiratory rates were only going up like one or two breaths, which didn’t make them like clinically like high respiratory rates, like clinically significant.

Emily Capodilupo: It was only significant in how it was compared to your baseline. And so that’s a case where like if I had gone to my doctor and they measured my respiratory rate, they would have said, this is a normal human respiratory rate, you know, between 12 and 20 breaths per minute, which is sort of normal. But like my baseline is about 14. So if it went up to 18, that’s a huge, huge rise for me, but it’s still technically clinically normal, so they would have completely missed that. But by having a wearable that’s like passively monitoring my respiratory rate every single night, you could see like something’s going on, and that can be a huge red flag that something’s going on with your respiratory system. Right. And of course, COVID-19 is a lower respiratory tract infection primarily. So it’s going to show up there. But we would expect to see similar things with somebody who had pneumonia or certain strains of the flu. And so these kind of like early warning signs that can show up in your vital signs before symptoms. You’re not going to have a fever yet. You’re not going to be complaining about not feeling well or have any other indication that you might have COVID. And so I think that’s like an example of where a wearable paired with a doctor can provide information that like a doctor in their office wouldn’t be able to provide alone.

Harry Glorikian: Well, I mean, I think, you know, if you took respiratory rate plus a slow change in temperature, right now you have two biomarkers that you can use to show something is physiologically off.

Emily Capodilupo: Yeah. What we were seeing was that respiratory rate was climbing before temperature was climbing, which was interesting.

Harry Glorikian: Interesting. Okay. You know, another story. It’s funny because I was talking to a friend of mine and he has A-fib [atrial fibrillation] and he knew he was going into A-fib and then he got together with his doctor and his doctor was actually digging into the data from the WHOOP to sort of see like when he was going into A-fib and sort of, you know, using the technology, because he wasn’t wearing a Holter monitor or anything like that. This, this sort of acted as a way for him to peer into when it started, how long it lasted and things like that. So I think when a doctor wants to, it’s interesting because some of these wearables like yours have that data available for them to, you know, interrogate.

Emily Capodilupo: Mm hmm. Yeah. And I think A-fib is such an interesting example there because, like, people who have paroxysmal A-fib can go into A-fib for just, like a couple of minutes a month. And so your typical like seven-day or 48-hour Holter monitor reading could easily miss it. But A-fib puts you at risk of all kinds of things like stroke that you might want to be treating, and so like having 24/7 data collection over months and months and months can give you a better picture versus I don’t really know too many people who are going to be willing to like or Holter monitor for a year.

Harry Glorikian: Yeah. So I mean, I’m going back to your 24/7 and the wearable and the fact that you’re driving all the power to the sensors, I mean, you guys collect, I think I saw the number, 50 to 100 megabytes of data per day, per user, which is a gigantic amount of data compared to maybe like a Fitbit or an Apple Watch. I mean. Why collect that much data? I mean, what do you do with it? I mean…

Emily Capodilupo: Yeah, great question. You know, we keep all of the data because it has tremendous research value in addition to being able to power the features that we’re providing today. You know, there’s all kinds of fascinating early research, you know, different things like the shape that your pulse makes. So if you look at not just how fast your heart is beating, but literally, you know what that raw, we called PPG, photoplethysmography signal, looks like, you can actually tell a lot about the health of a cardiovascular system. And we published a paper a couple of years ago now where we’re looking at age as a function of this like cardiovascular pulse shape. And we haven’t productized that research yet, but stuff that we’re exploring down the road and there’s just there’s so much, so much you can answer with large data sets that traditional academic research just hasn’t been able to answer because they haven’t had access to data like this. And so by keeping it all around, we’re able to do a lot of research and move the field forward as well as create really, really feature rich experiences for our members.

Harry Glorikian: Can I suggest, you know, custom consulting for guys like me who actually would love to dig into the data as as a service that that people would be willing to pay for. But correct me if I’m wrong — the WHOOP doesn’t really detect when I’m exercising. Right. I’ve got to tell it, no, I’m exercising.

Emily Capodilupo: We detect when you’re working out.

Harry Glorikian: Because it seems like it’s more accurate when I push the button first and it starts rather than wait for it to like if I’m about to start a weightlifting session, it’s more accurate when I push the button, then when I wait for it to tell I’m doing something.

Emily Capodilupo: Yeah. Well, with certain activities it’s hard to get the exact start times right. And different people have different attitudes about things like warm ups and downs and if they should be included. So if you do have a strong preference about whether or not you want those included, we do give people the opportunity to manually trim the bounds of their workouts or to just start and stop them manually. But we do detect any activity with a strain above an eight that lasts at least 15 minutes will get automatically detected.

Harry Glorikian: Okay. And by the way, I love the fact that you guys integrated with the Apple Watch because, like, because when I go on my treadmill, it automatically connects to the watch and then tracks the whole thing and then ports the info. That’s great. That is fantastic. As a as an opportunity. But, you know, how do you think about WHOOP versus any of the competitive technologies? And I’ll tell you why I say that when people say, well, what do you see is the difference? I’m like, you know, the Apple Watch is more of what what I think of as a data aggregation device in a sense, because it’s sort of taking all sorts of stuff. You know, the WHOOP I think of almost like a coach in a sense, as opposed to it’s pulling in data and pushing it out to different apps and I can do different things with it. So I don’t want to misrepresent how you might frame it, but that’s sort of how I think about it.

Emily Capodilupo: No, I think that’s totally spot on. I think that we have a very strong stance around not showing or generating data that we can’t tell you what to do with it. And so we really want to be like your coach or your trainer or at a minimum like your workout buddy kind of thing, where it’s somebody that or something you can kind of look to, to understand, you know, am I reaching my goals? What are the things that are helping and hurting me and sort of how do I then make changes to go forward? I think one of the biggest examples here is, we’ve been very much like countercultural in not counting steps and we’ve been asked a lot by our members, like, why don’t you count steps? It’s not actually that hard. It’s not because we can’t figure out how to do it. It’s that we actually don’t think that they’re valuable. Steps count the same if you run them or walk them. If you walk them upstairs or flat. You don’t get any steps if you swim for a mile and you certainly don’t get any steps if you’re wheelchair bound. And we didn’t like any of those constraints, they didn’t really make sense to us as a metric. And we also really didn’t like this kind of arbitrary, like everybody needs 10,000 steps. Well, is that true if I’m 90 versus 19, is that true f I ran a marathon yesterday, should I still be trying to get 10,000 steps today? Is it different if I’ve been sitting on the couch for three days? And so we came up with this metric of strain where instead of being an external metric, like steps are sort of something that you did and you can count them and it’s objective, we wanted an internal metric where it’s like, How did your body respond to that thing that you did and how much flow did you take as a function of what you’re capable of? And so sort of what strain does, it’s very much like in opposition to what steps does, is they’re internally normalized to reflect like if I ran versus walk to those steps, if I ran versus my brother ran and he’s more fit than I am, or if I do a two mile run this weekend and then I train a whole bunch and get more fit and then do the same two mile run six months from now, I should actually get a lower strain when I do it, when I’m more fit than I did when I got did it this weekend. Like all of a sudden, strain becomes this very rich thing because it has this, like, natural comparison where like a higher strain actually mean something objectively, both within and across people, than a lower strain does. Whereas that that’s not really true with steps. Right? I could walk fewer steps than you, but have done them up a mountain. And so I’ve actually put a lot more strain on my body than if I’d done the same number as you, but like flat pacing around my kitchen, eating snacks and making dinner or something like that.

Harry Glorikian: Yeah, well, actually there was an interesting paper that it was a sort of a study that brought in all sorts of studies to show that, you know, at an older age, you actually, you know, you need less steps, and it has a difference in mortality. And, you know, if you’re younger, then you want a higher level of steps. And, you know, so it was a good paper. I’ll actually I’ll send you the reference later. But you know, the interesting thing about strain is and this is the good part about the body and the bad part about the body, in a sense, is that it optimizes itself. Right. And so if you want to get the same strain goal and if you’re fit, you really have to…I mean, at some point, I’m like I look at if I had an incredible night, which is rare and it’s really in the green, I’m like, I’m never going to hit that. Like, I’m going to have to run ten miles to hit that, that goal. So, I mean, I try to like get out and lift that day and maybe get a run in, then get a walk in. And I’m still you know, when you can’t hit that high mark, if you’re actually in shape. When you’re not in shape, sort of, you can get there a little bit easier because your body is has optimized itself in a sense. Which is great, I guess. But when you’re when you’re holding yourself up to that number, you’re like, Oh, my God, I’m never going to hit that number.

Emily Capodilupo: Yeah. I mean, it’s super interesting how the human body works, right? There’s almost like this weird kindness in how we work where it’s like easier and more fun to make progress when you’re brand new and starting out and it’s harder to make progress the better you are.

Harry Glorikian: I mean, it’s an efficient machine. It has to optimize itself. Right. So, again, you were saying no display, no interface. All the information happens on the associated device, the phone. I mean, you mentioned some of the pros and cons, but are there any other that I haven’t asked or I know that at some point it pings me and says like. You need to connect because it’s been some time between connections. So is there an offloading time frame that it needs to…

Emily Capodilupo: No, it can store up to three days of data on the device itself.

Harry Glorikian: Oh, interesting. Okay.

Emily Capodilupo: Yeah. So if you like went camping for the weekend or something and didn’t have internet, we would just store the data locally and then transmit it all when you got back. But it tries to transmit the data more or less consistently, constantly throughout the day. What it’s pinging you about is not that you’re in any way in danger of losing the data, but just that you’re behind. And so you might be missing any kind of analysis or getting credit for your strains. We want to make sure you’re up to date so that if you want to look at your data from the day, you would have access to it.

Harry Glorikian: Here’s a question. Would it ever make sense to make a WHOOP app for the Apple Watch? Or is the device sort of inextricably linked to the app?

Emily Capodilupo: Yeah. I mean, there’s a lot of good reasons to think about something like that, right? You can make it a lot more affordable if you didn’t tie it to hardware. Right now, we believe that we have the best hardware on the market, but there’s sort of valid pushback that some people are willing to settle for something less than best in order to only wear one thing. And they want to wear their Apple Watch because they like the phone call notifications and the texting and email and all that kind of stuff. There’s a lot of great features that Apple has that we don’t. I’m certainly not trying to hate on the competitors at all. But I think like the way we kind of think about what we’ve done is like if Apple Watch does a lot of little things, you know, at like a relatively shallow depth, so it’s like a lot of coverage, we do a small subset of those things, but we do them very, very, very well. And so by not doing things like putting on a screen and letting you text and all of those things, we’re able to have all of the power of the device drive towards getting the most accurate signal data. And so we are sampling the heart rate more frequently than Apple is, and the device is more purpose built around optimizing both internally and externally for the sensors. So there’s even little things like electrical coupling on the circuit board. When you try and shove too much functionality into something small, they kind of like run into each other. And, you know, so we’re not trying to make room for a GPS chip or make room for a screen or like all of those things. And so it lets us lay out the hardware very specifically for this purpose. And so we believe that in data to support that, we’re getting more and more accurate like metric data.

[musical interlude]

Harry Glorikian: Let’s pause the conversation for a minute to talk about one small but important thing you can do, to help keep the podcast going. And that’s leave a rating and a review for the show on Apple Podcasts.

All you have to do is open the Apple Podcasts app on your smartphone, search for The Harry Glorikian Show, and scroll down to the Ratings & Reviews section. Tap the stars to rate the show, and then tap the link that says Write a Review to leave your comments.

It’ll only take a minute, but you’ll be doing a lot to help other listeners discover the show.

And one more thing. If you like the interviews we do here on the show I know you’ll like my new book, The Future You: How Artificial Intelligence Can Help You Get Healthier, Stress Less, and Live Longer.

It’s a friendly and accessible tour of all the ways today’s information technologies are helping us diagnose diseases faster, treat them more precisely, and create personalized diet and exercise programs to prevent them in the first place.

The book is now available in print and ebook formats. Just go to Amazon or Barnes & Noble and search for The Future You by Harry Glorikian.

And now, back to the show.

[musical interlude]

Harry Glorikian: So switching to sort of business model, because you sort of touched on that, is like it’s a subscription model. You don’t buy the device. If I’m not mistaken. The service starts at say 30 bucks a month and the package actually includes the WHOOP band. They’ll just ship it to you like I’m wearing mine. Right. And so what was the rationale behind subscription versus just selling the device. If you have insight into, how did they pick 30 bucks? You know, I just wonder, like, you know, did they, is that something you guys felt reaches the broadest market sort of thing?

Emily Capodilupo: Yeah, pretty much. So when we actually first launched, it was sold more like a traditional hardware product. So it was $500, one time fee, sort of use it as long as you want. And then we switched over to the subscription model in 2018. A nd we chose the price of $30. It was sort of designed to make the product accessible and lower the barrier of entry. $500 up front is a lot of money, especially for younger athletes. We want to make sure that people in college could afford it and stuff like that. And so we found just by market testing, that $30 was an approachable price point. And so after a couple of different market tests, that was what we landed with and more or less where we’ve been. We occasionally discount it and different things like that, and you can get a lower rate if you commit to more months upfront.

Harry Glorikian: Yeah, I think I signed up for the maximum, which then brought it down to I think it was $18. Yeah. So here’s a, you know, because this show is, you know, supposed to focus on AI and health care and things like that, I’m just sort of imagining in the back of my mind with that much data, you really have the opportunity to build some really cool analytics on top of it. You know, what role, if any, like does machine learning or other forms of AI play in you know how you analyze the data and then how do you, do you actually use that to personalize it back to the individual using it.

Emily Capodilupo: Yeah, I mean, that’s pretty much all my team is doing is machine learning. No, it plays a huge role in what we’re doing, from like very traditional ML approaches, so like if you think about how we’re doing our sleep staging, we have polysomnography is like the gold standard for getting sleep truth data. So that’s like the stages when we know we’re in REM sleep or slow-wave sleep. So we sent thousands and thousands of people into a clinical sleep lab with two straps on and they underwent a clinical sleep study. And then we took all of the data from the sleep study, lined it up with the WHOOP data, and then used all kinds of different traditional ML approaches in order to figure out how to get from a strap the same sleep staging information that we’re able to get from this gold standard approach. Obviously the sort of gold standard sleep study uses a lot of sensors that we don’t have right things. EEGs, which you need to be on someone’s head to use. You can’t get EEG from the wrist. EOGs, which you have to measure eye movement. So you need a little sensor there. And then we were able to find good proxies from the data that we can get at the wrist for all of those different signals and reconstruct the same sleep stage information.

Emily Capodilupo: So that’s a super fun ML problem. We also do things like when we detect a workout, we can figure out what, which sport or exercise modality you’re using. And so the ability to classify those workouts is kind of again like a traditional ML like time series classification problem where you can tell the difference just from the heart rate and accelerometer signals. Are you doing basketball or CrossFit or running or anything like that? And then so those are kind of more traditional ML approaches. And then we’ve also done a lot around trying to understand behavioral impacts and how your body responds to different things. And then we’re doing things like much, much more personalized. So we have a feature called The Journal where every day you fill out this little diary and you answer a bunch of questions about what you’ve done in the last 24 hours and can self report things like when you were eating, if you did different like kind of wellness activities like, meditate, journal. You know.

Harry Glorikian: How much alcohol you had. I always wonder, like how honestly somebody answers that question.

Emily Capodilupo: Any of those kinds of things. And then we look at the sort of signals in your data and try and separate out which of the things are helping you, which are hurting you, so that we can then recommend the things that are good for you, and for the things that are less good for you, maybe help you quantify the cost of those things that you can deploy them strategically. We certainly don’t expect everybody to become like a teetotaller and never drink again, even though we’re going to tell you it’s bad for you, because it’s pretty much always what shows up in the data. But we do want to help people make those informed decisions because a lot of people think like, Oh, I can have two drinks and it won’t affect me tomorrow. And like, okay, here’s the effect. And if tomorrow’s not that important, go for it. And you have that really important meeting tomorrow, maybe don’t. Y rou know, we’re not trying to kill all the fun by any means, but we do want to make sure that people are empowered by data to know understand what they’re doing to their body and then make decisions accordingly.

Harry Glorikian: So I’m throwing in sort of like something important to me, right? Which is, you know, I have sleep apnea. Right. And it’s funny because my wife diagnosed me, but then, you know, all the devices at some point, my Apple Watch actually asked me once, you know, have you ever been diagnosed with sleep apnea, which was interesting. But I’ve noticed like, the recovery number, if don’t wear my CPAP, my recovery number tends to be much higher than if I do wear my CPAP. And I always wonder, does the positive air pressure cause a difference in how much your heart actually rests or not? Because it is pushing, it is positive air pressure on you all the time. So even in between apneas, you don’t really maybe not rest as much. And I’m wondering if you have any insight on that.

Emily Capodilupo: Yeah, we, we haven’t specifically dug into why, but we have seen that as an unexpected pattern. You’re not the only person to report that. It’s on the to do list to better understand what’s going on there. I think your theory is a valid one. We haven’t verified or ruled it out yet, but I think there’s a lot to be learned there. And I think one of the things that’s exciting about the data that we’re collecting is that if you wear a CPAP is one of the things you can report in our journals. We do have a tremendous amount of data on that and therefore the ability to kind of tease that apart and get insights that haven’t been made available yet by traditional academic research.

Harry Glorikian: Oh, I didn’t know I could add CPAP in there. I have to go back and and check. But yeah, because my strain score ends up, my recovery score ends up lower. So it’s like, you know, then of course, I always exceed on the strain side because I’m going to go work out the next day. And you know, it is what it is. But the other thing that you guys offer is like WHOOP for teams. And I don’t know if you mean sports teams. You mean organizations. I’m not 100% sure because obviously I don’t use that. I’m using it as an individual. Can you explain the additional value that provides when a group of people are using it together?

Emily Capodilupo: Yeah. So all the above, we do it corporate teams as well as athletic teams, and there’s a couple of different layers of the added value. So sometimes it’s just accountability. I’m on a team with my family and it’s just kind of fun, make fun of each other when our recovery scores are poor and, you know, cheer each other on when we have particularly good strain scores. And, you know, there’s a lot of data to support that when you have a workout buddy or an accountability buddy or anything like that, that you tend to stick with things longer. And so creating just like a really friendly way for people to compete and cheer for each other just helps with the accountability and motivation keeping people on track. And deeper and more importantly, we do have a lot of people who create teams around different kinds of research initiatives or trying to understand a certain life stage. Like we create teams for people based on the month that their babies are due. So pregnant women can join a team of all the women on WHOOP who are expecting a baby in June 2022 can join this team together and pregnancy is this like very foreign weird moment in your body where everything’s changing all the time and it just creates, like, a way for people to connect and be, like, this weird thing that’s happening to me, is it normal? Like, who else is sleeping funny? And I think it’s just very comforting to know that, like, all these weird things happening to your body aren’t so weird. And then with like the sports teams and different things like that, what we’re seeing is that the coaches are using the information to make better training or like decisions because now they actually have information that they didn’t have access to before.

Emily Capodilupo: So we’ve done a lot of work with different like collegiate programs and professional programs where they do things like if you’re red, they will have you do a lighter version of the practice or skip a section of the practice in order to give your body a chance to recover. And if you’re green, they might have you push a little bit harder. And so by modulating the training to where your body is today, we’ve actually shown in a project we completed a little over two years ago that you can reduce injury without reducing performance gains over the course of like an eight week training period. And so by reducing your training, when you’re red, so your recovery score is below 33%, you actually like you will reduce injury without reducing performance gains. We’ve shown this. And so there’s like literally zero value for those coaches to like push the athletes to complete the program or the day’s rtraining. And so we’ve seen a lot of coaches make those different training plans as well as game day decisions about who should start. You know, somebody might be your best player ordinarily, but if they’re red, they’re not all that primed on game day to perform. And so being able to make those kinds of different decisions. And then on the corporate side, people have used it in order to triage different access to supportive resources. So we’ve seen people offer like breaks to people who have been red for a number of different days in a row or things like that suggest that somebody might be burning out or overwhelmed or something like that.

Harry Glorikian: Okay, so. Everywhere it states that it is not a medical device, is not intended to diagnose, monitor any disease or medical condition. Right. What’s the line in your mind between, say, a fitness monitor and a medical device, because I think I always think that line is getting….because you guys and others like you guys have so much data, the level of insight that I’ve seen when I’ve gone into some of these is crazy. So. What what is that line in your mind?

Emily Capodilupo: Yeah. I mean, I think that there’s you know, it’s always been the case that technology moves faster than the law. And so, like, you know, I think a lot of these things are going to shift as the technology is going to force them to shift. But, you know, like you said, we have a lot of data that’s quite similar. The official line is what the FDA says is the line. And the FDA has carved out this like space that they’ve you know, they’ve called this wellness devices. They’ve sort of reserved the right to change their mind at any time, and we very much expect them to. But WHOOP falls into their definition of what a wellness device is, not a medical device, which is why we can say things like, this is your heart rate, but we can’t say, because then you would cross into a medical device, like “Your heart rate is healthy, your heart rate is unhealthy,” right? You can’t give those kinds of any kind of diagnoses or any kind of, like, you will prevent a heart attack if you do these things or something like that. So we have to keep the recommendations a bit more general, a little bit more vague in order to not cross over into that regulated health space. One of the things that we’re seeing that’s interesting, is that there’s been a movement in wearables to get these like SAMD clearances, Software as a Medical Device, where pieces of wearables need different features or different algorithms do end up going through an FDA process and getting clearance to make certain claims in different settings.

Emily Capodilupo: And I think that that’s going to really accelerate over the next couple of years. These are very long processes, and then the lines are going to get more and more blurry because you’re going to have this like hybrid consumer medical device, which is something that until a couple of years ago we really didn’t have. There was like step counters and GPS watches and they were over here and then there was like medical stuff that didn’t look cool and wasn’t comfortable or easy to use and was very, very expensive. And it was all over here. And now we’re seeing them kind of come into the middle where more and more the medical stuff cares about being like all the human factors like that’s comfortable to use and that people want to wear it and they can get good compliance. And the wellness devices are finding more and more applications for their data in the health care space. So I think a lot of it’s going to come down to what doctors end up getting trained on. If they’re willing to look at this data, if they have any clue how to use it, sort of by being in the medical world and science training their whole lives, a lot of them just don’t have the education and training to understand big data and to understand technology in that way. So they’re not being trained on how to make use of the data or how to apply it. And I think that that’s something that might change in the next couple of decades.

Harry Glorikian: Well, it’s interesting, right, because I always tell people I’m like, this is a medical device. Like I you know, I mean, you know, you may think it’s not, but it really has certain capabilities that allow it to get FDA clearance in a particular area. Right. And they’re picking their space one by one. But the amount of data that you guys pick up on all of these devices, I mean, you know, we’ve seen atrial fibrillation. I’m sure that tachycardia shows up on there. You know, there’s different things that they, because it’s 24/7, it’s looking, right and it’s monitoring and it’s got multiple sensors which you can now cross-correlate. There’s so much insight that comes from this that I would almost like love to encourage the companies to think about moving down this road because I think it would be so helpful to patients. But, you know, jumping to a different thing. So. How do you guys define success for WHOOP? If you hit all your product and sales goals and for the next, say, 2 to 5 years, what does success look like for the organization?

Emily Capodilupo: Yeah. I mean, I’ll let the finance team worry about the sales goals and things, but I mean, for me in my team, like what success really comes down to is like, can we help people make actually better decisions? I think like a lot of the first generation of wearables, like it was this stream of fun facts. And we’re all obsessed with ourselves, right? Like humans are sort of naturally narcissists, at least to a certain extent. And so it’s like fun to be like, ooh, I slept for 7 hours or like, ooh, I ran a mile. But it’s like kind of you maybe already knew that, right? And I think, like, what we’re trying to do and like where we see a lot of success is, can we tell you something that you don’t know? And can we convince you that you should do something about it? And then can we make you, like, realize, like, oh, wow, this, like, incredible thing happened and I feel so much better. And the features that we get the most excited about are like the sort of user stories are not, like, “Wow, it’s so much fun to see my sleep data” or like, “This was fun.” But like when we released our paper showing that this respiratory rate spike sort of predicted or often preceded COVID symptom onset and therefore COVID infection, the paper came out like right before Thanksgiving and we saw so many people tell us that like because they had a respiratory rate spike, they didn’t go home for Thanksgiving or they didn’t travel and then like they tested positive a few days later and they were like, my grandma was at Thanksgiving or like my uncle who’s in his eighties or stuff like that.

Emily Capodilupo: And you know, those kind of moments where it’s like, we educated you, we showed you this vital sign that like, you never would have felt anything. You didn’t know you were sick, you weren’t feeling bad. It’s not like you went to go get a test because you weren’t feeling good, like you just saw this in your WHOOP data and you’re like, You know what? I’m going to stay home and not risk like seeing grandma because WHOOP said so, right? And then like, who knows how many COVID infections didn’t happen and like what kind of role we played there. And like, it was probably like the most meaningful thing we did that year. And we did a lot of other cool stuff, but to think that by helping people notice that pattern, potentially they saved a relative’s life and all the like crappy things that would happen if you thought you were responsible for killing your grandma and how much that ruins your own life as well? I think like we just get really excited about that. And one of the features that we released is last year was we were looking at how your reproductive hormones is part of your menstrual cycle affect your ability to respond to training. And I was an athlete my whole life. I was a gymnast, like before I could walk, and like nobody asked me a single time when my last period was or anything like that. That was just totally not part of like the coach-athlete relationship. But we know that like your ability to put on muscle and your ability to recover from training is totally different during the follicular phase, the first half of your menstrual cycle, than it is during the luteal phase, which is the second half. And if we modulate your training so that you’re training more during the first half of the cycle than the second half, you can way more efficiently build muscle and strength, have fewer injuries, make more efficient gains. And if we now we do coach, in our product, women to do this, and we’ve gotten this incredible feedback of like people saying they feel so much better and like they’re, well, you know, their training is going more smoothly and they feel like their body so much less random, it feels more predictable and they kind of understand what’s going on. Nobody ever told them that reproductive hormones were relevant beyond their role in reproduction, but they actually affect everything we do. Like when progesterone is elevated in the back half of our menstrual cycle during the luteal phase, we sweat more and we lose a lot of salt by doing that. And so we need to eat more salty foods and we need to be more careful about hydrating, which is really important if you’re an athlete, but nobody’s telling us this. And so like we can connect these by looking at big data because we are tracking your menstrual cycle around the clock or around the month.

Emily Capodilupo: We can put that into the product and then we see people are making better training decisions, understanding their body, feeling like things are less random. Right. And that’s so empowering. And I think like female athletes in particular have been so underrepresented in research. There’s a paper that came out eight months ago that said that just 6% of athletic performance research focused on women, 6%. And it was looking at all research between 2014 and 2020. And it was trending down, not up. So it was worse in like 2018, ’19 and ’20 than it had been like earlier in the twenty-teens. And so it’s like completely neglected. And there is all this data that like wearables and WHOOP are sitting on and we’re able to create features around that and just help people understand their bodies in a way that nobody else is doing right now. And so those are the features that, like I really define as like big successes. If we made our sleep staging accuracy 1% more accurate or we caught one more workout, like those are obviously like from a pure data science perspective, they can feel like wins. But what we really care about is like, am I helping you, cheesily going back to our mission, am I helping you unlock your performance in some way by helping you understand your body and making a better decision? Like, are you better off for having been on WHOOP? That’s what, internally, those are the KPIs that we track the most closely.

Harry Glorikian: Yeah. And I mean I would encourage you as well as all the other companies to, you know, peer reviewed papers, get them out there. Right. I mean, just when I search the space or peer reviewed journals for things utilizing the technologies, I mean, there’s not a whole lot out there. And then the other thing is, is sometimes I read the devices they’re using, I’m like, whoa, what is that? I’ve never heard of that device. And if I haven’t heard about it, it must be on the fringe sort of thing. So I would highly encourage it because, you know, people like me would love to be looking at that sort of data. Because I’m constantly investing in the space, constantly working with the different technologies, you know, constantly talking to people through the podcast or writing a book, you know. So that information is incredibly useful to someone like me as, as, as well as the average person. So if you could send a message back through time to yourself in 2013 when you joined the company, you know. What would you say? What have you learned about the wearables and fitness market that you know you wish you knew then?

Emily Capodilupo: Oh, what a fun question. You know, I think, like. It’s hard to know what I wish I knew earlier because like in so many ways and I feel so lucky that this is true, like the vision that Will pitched me on when I met him, like when he was like, “Come join WHOOP, this is why it’s super cool,” is exactly what we’re doing. And so, like, I did trust him. I guess my message in a lot of ways would be trust him that like this is for real. I think the space has been so exciting and just there’s so much opportunity. I came from doing academic sleep research and I would work on these papers where we had like 14 subjects and it was like, “Oh, that’s a, that’s a good size sleep study. Like that’ll get into a good journal.” And everyone was like excited. And then it’s like, you know, I just, I’m working on a paper right now and we have 300,000 people’s data in it. We’re looking at like a year of data at a time. So we’ve got just like millions and millions of sleeps and workouts in this data set that we’re combing through. When we did this project, which was published in the British Medical Journal last year, where we were looking at the menstrual cycle phases and how they affected your training, we looked at 14,000 menstrual cycles, like just the orders of magnitude more data than what you can do in traditional academic research. And that’s what I got really excited about. It’s why I became a data scientist because I realized that like the most interesting questions that there are to answer about how humans work are going to require larger datasets than we’ve had access to before.

Harry Glorikian: So I’m putting in a plug for sleep apnea, man, if you get a chance, I’d love to see a study on that one.

Emily Capodilupo: No, sleep apnea, it’s definitely on the list. About 80% of sleep apnea is believed to be undiagnosed. And it does have tremendous effects on long term health when it goes undiagnosed, especially in later stages. And so anything we can do around helping people realize that they might have sleep apnea and then helping them treat it once they do and better understand the disease progression. And all of that has a huge quality of life implications down the road.

Harry Glorikian: I will happily volunteer. So great to speak to you. Very insightful discussion. I’m going to tell my wife about the whole menstrual cycle thing and working out and this is exactly why she eats salty food like at certain times. But this is great. I’m so glad to have you on the show and I look forward to seeing the progress of the company and the technology.

Emily Capodilupo: Awesome. Well, thank you so much for having me. This is such a fun conversation.

Harry Glorikian: Thank you.

Harry Glorikian: That’s it for this week’s episode.

You can find a full transcript of this episode as well as the full archive of episodes of The Harry Glorikian Show and MoneyBall Medicine at our website. Just go to glorikian.com and click on the tab Podcasts.

I’d like to thank our listeners for boosting The Harry Glorikian Show into the top three percent of global podcasts.

If you want to be sure to get every new episode of the show automatically, be sure to open Apple Podcasts or your favorite podcast player and hit follow or subscribe.

Don’t forget to leave us a rating and review on Apple Podcasts.

And we always love to hear from listeners on Twitter, where you can find me at hglorikian.

Thanks for listening, stay healthy, and be sure to tune in two weeks from now for our next interview.