clock menu more-arrow no yes mobile

Filed under:

Full transcript: Peloton CTO Yony Feng on Recode Decode

Meet one of the companies trying to shake up how we exercise.

Amelia Krales for Recode

On a recent episode of Recode Decode, Peloton CTO Yony Feng spoke with Recode’s Kara Swisher about how the company distributes virtual spinning classes to its exercise bikes around the world.

You can read some of the highlights from the interview at that link, or listen to it in the audio player above. Below, we’ve posted a lightly edited complete transcript of their conversation.

If you like this, be sure to subscribe to Recode Decode on iTunes, Google Play Music, TuneIn and Stitcher.

Transcript by Maya Goldberg-Safir.

Kara Swisher: So I was in New York and I got an email from your PR person and I've always been interested in Peloton, so I was very happy I was able to talk to you here, because it’s sort of in a genre, an interesting genre of the Internet of Things, but in a wholly different way. So why don’t you explain what you're doing and how it’s different in the exercise space.

Yony Feng: Certainly. The idea of Peloton is that you're taking a high-energy group fitness class, in this case on an indoor cycle and trying to recreate that redistribute, ecosystem, energy that you're generating in a classroom setting in a geographically distributed ...

People are — everywhere — taking these classes?

Virtually taking this class everywhere. And kinda transferring that same level of energy in a single classroom but to, you know, hundreds or thousands of people who are taking it in remote locations, dispersed but at the same time they're all brought together.

So how is that so much different than — there used to be shows on television, exercise shows that people followed along with or — you know, you're not as old as I am but Jane Fonda had a series of videos that were super popular workout videos that people did at home. What is the difference in your perspective of what you're doing?

So in the traditional broadcast where you're looking at exercise routines on the television screen, it’s more one directional. You're looking at the instructions.

Or they have them on Xbox, things like that.

Or p90x, the routine is high energy, high interval. In our case, it’s that but so much more because you are looking at other people who are not just those actors, actresses or fitness professionals on a screen, you're also looking at other people just like yourself and what they're doing: Their stats, their metrics from their bikes and how you rank up against them. So it’s kind of merging the world of the social media mass of more people ...

So you're in a class essentially, but not; you're at home doing this. So explain what you do, you buy the bike for — what is it? — almost $2000.

Correct. $1,995.

Okay, $2000, and then you pay a monthly fee like you would to a gym for the videos, and they're from where? Explain where they're from. So you have a pad on the bike, correct? Like a wireless pad?

Yes, there's a console attached to the bike that is gathering the metrics from the bike itself. You're paying $38 per month of subscription fee and we are generating the content, we're live broadcasting the content from our New York indoor cycling studio, production studio.

You have a studio that people go to?

We have a production studio. Think of it as "Saturday Night Live," but there are actual participants in the studio who are taking the classes as their daily routine. We feel that's a good way to transfer the same amount of energy. By having live participants in the livestreams — we have about 12 livestreams per day happening — and this is generated from the studio and broadcasted to everyone who owns a bike at home.

So they wanna do it live, or they watch it taped?

Both! So they can, if it fits their schedule, the whole entire mantra is that you work out on your own schedule. So if it — we have 12 classes per day — if it just so happens to fit your schedule, that's great, take it live. If it doesn't, you can take it on demand as well, we have some of the classes that are put on demand so you can select from one that maybe was aired this morning at 9 o'clock ,so you wanna take it again at 9 pm tonight cause you weren't available at 9am this morning.

And you want that particular class.

Exactly. You want that instructor, you want that playlist, you want that - you know, you knew your friend took it at 9 am and you wanna take that same class at 9 pm.

Now this is an expensive thing, $2000 and $39 a month. Gimme the background of thinking of doing this. It’s been around since 2012, right?

2012, yes. And at that time we were struggling on thinking about the pricing for this bike and we always stuck to our core business model, which is subscription. We want to focus on the subscription, which is vastly different than your traditional gym equipment manufacturer which is trying to make a margin on the first sale. And in my mind, as someone like that, I don’t really want people to use it too much, because I'm kinda guesstimating a little bit because if you do, there's a chance of breakage on normal use. Because for us, we're making this bike and we want people to use it because the more you use it, the more value you get out of it and the longer you pay a subscription.

Subscription fee, so you wanna make something exciting for them. So this bike gets delivered to the home, but it’s still a high price, who are you aiming at? Is it millennials, is it rich guys or what? What’s the — people who don’t wanna go to SoulCycle?

So our — what we're targeting is, first of all the bike, you look at the price point of the bike, our goal is to make the barrier to entry low. If you look at the quality of the bike and the features on the bike, you look at other types of bike that have the similar features such as magnetic resistance and a [silent] belt, there's a certain price point for these features with the bike. Of course we coulda just decided not to have these features and have, you know, a $300 bike, but we believe that if you do that, the user experience of riding on the bike, it’s just so degraded that you’ll just use it once.

You don’t enjoy it.

You don’t enjoy it, it’s not an accurate simulation of outdoor cycling and you'll get on this bike and you'll say oh, this is like a toy. And okay, I spent $300 on this bike and I'm probably never ever gonna use it again, so that would be detriment to our user.

So you're trying to get people to use this online experience, essentially, to do it. So talk a little bit about what’s the technology behind it. It’s a pad, it’s connected to either live or on-demand streams, just like you'd be watching Netflix or anything else.

So there are a few pieces of electronics on this bike. There's the — we have, I believe we have a total of two sensors, on the bike. One is for collecting RPM and also another one for collecting the resistance and through that we calculate the power that you're generating on the bike. Those metrics are sent to our console that is running our customized operating system and it is able to then aggregate that metric and data as a hub and send it to our cloud platform for redistribution to everyone who is taking this class at this same.

And how many people take the class at the same?

Let's see, uh, at the same time our platform has thousands of people taking it at the same time, different classes, of course. I think one of our largest live rides, which usually happens during the winter times when the weather is not ideal for cycling, we have more than 500 people taking a live class at the same time.

So do people then, as a participant — I assume you hire those people as participants, or maybe you just offer them classes, it’s in a studio —

Actually, those are just regular people coming in, and they're actually paying for classes.

Okay, like a SoulCycle. And you talk to the people at home, do they need that? Like you suddenly say, Kara keep up , or something like that.

We found that — initially we weren't sure if that was a good thing or not, will people be freaked out by it, will people just ignore that or what will people do? Will they actually turn the volume down and watch a TV show? What we found over time is that — and the way we find out is we have a very active Facebook group of Peloton riders — they're kinda taking live rides and hoping their name gets called out, and people are kinda like, "Is my name gonna be mentioned? this live ride I'm gonna take?" And what they do, they kinda post it on FB and share that excitement with other people and of course our coaches are also on the FB page and they're communicating, they're sharing their experiences — "Really loved that last song, that last song really gave me the energy to push and make me break my own personal record" —

And then everyone sees those statistics too, correct? But not necessarily your name ...


So only your username, which could be anything, it’s an avatar username, which is seen public.

It’s an interesting thing because I, for example, I don’t go to SoulCycle because I live in the Castro and there's all these incredibly fit gay men there that I don’t wanna keep up with, and I don’t want them to insult me, either. But you sit in the back and you have a little too much self awareness of yourself when you're in a class, but some people love them, people love sort of the group aspect of it. So I'm wondering how it translates when you're by yourself, the idea of — cause they get kinda cultish with their obsession with the group while they’re physically there — but how do you manifest that when you're alone in an online setting?

Yep, so that was always the question in our quest to create this user experience and product: Is it actually gonna translate well from a one-classroom setting to you being by yourself in your home gym or your living room or bedroom? And what we found is that the leaderboard and the interaction with the instructor, what they're telling you and talking to you, that combined gives you enough of a translation of the energy that's happening along with the broader class. You see people in the background also kinda following the instructions from the instructor in the front, it gives you a sense that you're part of something bigger.

That you're there.

And the leaderboard is doing that as well, you see people's number climbing and you're just like oh my gosh, this person is beating me, or if you don’t care about the leaderboard, you can just see other people's names shifting up and down. The one special thing is their location, actually, you see their username and you see whatever people put as their location — I'm from Connecticut, I'm from Florida, I'm from California, and you're like, wow this is a truly distributed experience.

Let's talk about where this goes because this is sort of the beginning of the idea of how exercise changes. I mean, we started with broadcast exercise on television and then video tapes and then it was on a lot of platforms, all the kinda different things, and now it’s moving to this. Talk about where the technology of this is. It’s an Internet of Things kind of play, where that's going.

Certainly, if you look at where that's alluded to earlier in the show, the traditional fitness classes, whatever it might be — broadcasting with the one-way communication — engaging people, but this is adding another layer of real-time data while people are participating. If you look at other apps such as Periscope, Facebook Live, people are starting to realize that the real engagement is actually in Live, because there's the sense that people are there at that moment in time and the shared experiences of everyone at that moment in time, looking at something and doing something simultaneously. [It’s] about what they're absorbing in terms of content or physical activity, is a shared experience, it doesn't matter if two people are 2000 miles away or right next to each other. So the Internet of Things, the metric connection really gives you the data platform for what to share during those live moments. What’s important is to really keep in mind what engagement and what experience we're trying to portray in an indoor cycling class. What’s important? Is it heart rate, is it the power you're generating, or is it actually how much you're sweating and how much you're breathing? what do other people care about?

So there's been a lot of these, the Fitbits, the apps, all the things where people are taking [measurements], digital means to share or do teamwork to do better. Where does this go, this concept of exercise with an added digital part? I mean, obviously, heart-rate monitors have boomed, all kinds of watches, Apple has a watch which hasn't done that well. And I think people were surprised how not well it’s been doing. Where do you see the industry going? Where is it headed?

I think the industry started by kind of using a shotgun approach, saying we're gonna track everything that anyone does with everything, right, and I think that as the sensors and Internet of Things, platform, mature, the barrier to entry becomes lower. So now you don’t have to be a tech-focused engineering company with 5000 engineers in order to produce something that can track your movements. As something that's become monetized more, what’s species — that have traditionally focused on specific experiences, such as Peloton, we focus on the indoor cycling experience, then we start looking at the sensors and how they can enhance the existing proven form of exercise, such as indoor cycling, such as potentially yoga or strength training or stretching. How do these sensors actually help focused periods of time of your workout?

Do you imagine sensors all over people, like they put them all on their bodies to do that, like yoga, give an example of yoga. You guys are working on other things besides cycling, so strength training you have to sell them some sort of weight situation, or yoga maybe a mat.

Right. There are many many ways. For example, yoga: There are clothing out there that actually tracks the contraction of your muscle and the usage of your muscle, how much the muscles are actually stretching. There are bands that are tracking your perspiration on the skin, and of course yoga heart rate is important as well so maybe you want to maintain in a reasonable heart rate zone. In strength training, clothing becomes wearable to another level, which is fabric, tracking the electric pulses in your muscle. Those can also be very much informative but also critical to doing maybe what you're trying to achieve as part of the workout.

So you could in that case, on a bike, the thing is right in front of you. Like, you can look at a thing. If you're doing yoga, you'd have to look at a screen presumably, so it would be a larger television experience.

So one idea is a television experience such as Apple TV which supports actually your second-screen experience. You have the video there and you have the metrics part, either on your iPad or actually on the right side of the screen of your television, but at the same time, Peloton actually manufactures its own console as well so I mean we chose the size of a console, there's nothing that stops us from making a larger console as a home entertainment exercise console. So there are all these ideas and possibilities that we already have the experience and resources of doing already actually.

So if we call it Nintendo or one of them, the Wii platform tried that with things that you jumped on and you know there was a platform, I had some, there were all kinds of things. It didn’t really work out, it didn’t continue. What was the fault there? Because I recall a platform that you'd have to jump on and off.

Yes, absolutely. I think it was cardio step platform where you know you're stepping up, out, down.

And Kinect was a thing cause you were watching your body movement and stuff like that.

I think Microsoft Fitness just started to wind down I think two weeks ago. So I think a lot of it is they took it, they looked at what is out there and they looked at what they have, and they thought about the experience within the confines of the existing platform they have, which is for Xbox, the Xbox platform, and for Nintendo, the Nintendo platform — and Nintendo actually, I believe it was Nintendo, tried to create an indoor cycling all-in-one, with a console, 25, 30 years ago? With the CrT2. So for us, I think the difference for us is that we have our co-founders John, Tom, Graham and myself, have all been avid users of the indoor cycling experience and also other HIT experience. We focus on the experience first and foremost, we forget about the technology and we just look at the experience and see why people love this experience.

So these things were add-ons, they were playing games, and also playing tennis — and my kids played it for a while, I have to say, an then just didn’t!

There's a novelty and then, I think, right, there's a novelty part of it.

You held the thing in your hand.

The novelty, and then once it wears off, the content is stagnant, it’s always the same style. And I think for Peloton, it’s not an indoor cycling experience on top of the gaming console or a rider on top of. This is a core experience and we have an entire studio, broadcast studio dedicated to producing this content, you know, we have lighting, we have sound, the instructors are the best amongst the class for this type of content, so every piece of the experience and the ecosystem has to be focused on being the best in class for cycling, for exercise, for fitness.

But how do you prevent it from being a fad, because exercise is plagued by fads. Either there's new kinds of exercise places or there's new — suddenly it’s a Zumba or there's this or there's that, suddenly it’s boxing and stuff like that, it has this issue.

So for us, you have to define what the experience is. You cannot just listen to every suggestion there is about what a finished routine is. We have our defining program, we have our defining training program for our content, we have defined what this experience, people at home are committing to for 45 minutes, the routine, and we take it very seriously how we interact with our users and what type of feedback they are providing to continue to make sure that we're not diluting our experience. We're not, you know, veering off too much.

Alright, but you have to get new stuff, yoga and things like that.

We don’t have to get new stuff, I think the experience on the bike is the experience on the bike. It’s not gonna be diluted by oh, you know, for the next two minutes.

You do have a popular, you have a studio in New York, correct?

We have a studio in New York and that is our main studio. Our plan is not to expand to more number studios, we're only one broadcast studio and maybe in the future if we go international we may wanna create one or two more for timezone reasons, but mainly ...

Are you bringing it to California? A lot of your users seem to be, a lot of the VCs seem to like Peloton.

Yeah, certainly, I would love to have a studio in California myself but that is not in the plans right now, we want to use our studio here to really focus on the quality.

So you're not getting into the SoulCycle business.

We are not getting into the SoulCycle business.

But they are getting into yours apparently, they've discussed it.

Yes, that is something.

The idea of selling SoulCycle bikes and then broadcasting from their studios.

Yes, that has been mentioned many times. I have not personally seen any concrete demonstration or product demos yet.

It’s not unusual that they would try to do something like that to expand their businesses.

Yes. We've been anticipating actually a —

Copycat —

For some time and it’s actually curious, why it kinda hasn't happened.

Why hasn't it happened?

Well, I think it’s a very good question. If we look at what we're doing right, we're on the manufacturing side of gym equipment, the Peloton bike is innovative in design and engineering, and we are also producing content, so imagine you're HBO, Netflix, developing original content for this platform. And we have the data platform, which is the real time leaderboard, the achievements, the user profile, the cloud platform that is powering the data portion of the real time and offline experience for users. So these three pillars or verticals are traditionally very separate until recent years and for us we're one of the first specifically focused on the indoor cycling.

So what else can you do with this data? Could people wear Peloton, you know, watches with that kinda stuff. Or what’s the, you want more information from people?

Yeah, so our direction is not to be a fitness app of indoor cycling! We are not really interested in what your entire 24-hour, what your sleep pattern is, when you're sleeping after taking a Peloton.

What do you think of all those, they sort of go in and out, they're like bread machines.

Yeah! I think quantified —

Quantified self.

Yeah, movement which is measuring every single bit of information about your body, about yourself and then providing analytics about your way of life, your lifestyle. It’s super useful. I love it myself. I think for Peloton, it’s a drastically different direction than that, it is not a health tracking platform that is tracking every moment of your life. It is to provide you with an active 45-minute, focused engagement in that sense. And our definition of offline experience is your connection to instructors and your connection to other people who have taken this class.

So where does that go? Can you imagine a VR experience? Or an AR experience?

So me personally, before the launch of Peloton bike, I bought an Oculus VR.

Course you did.

[laughing] I had to! And I actually emailed Oculus VR a few times saying, is there something we can do together? I've always been interested in this area. Unfortunately for myself actually, I can use a maximum of probably five minutes of VR before I get nauseous, and Gear VR, the newest Oculus VR product.

Supposedly they've improved that.

Right, and of course, the future Playstation, Xbox VR experience are closely monitored by me personally and other members of our engineering team. We actually created a few demo classes on the Gear VR platform.

So people would be in their home, get on their bike, and then they would put on a VR thing, looking completely ridiculous, and then they would be in the class.

Right, so kind of you're sweating on the Gear VR, fogging up the lens a little bit but the experience — it really felt real! You're inside.

That's the point of VR, I think. [laughter] I believe that's the definition.

It was a great translation, actually. But I think there are a few maturity cycles needed for the hardware and the delivering platform before we can really start thinking about having people put this on for 45 minutes, 90 minutes. To have a very high energy [experience], your heart rate is 150 to 170.

It’s a big device to put on your head, to be exercising in.


But do you see a time, is that where it’s going, do you imagine? Or can people just go to classes, because there is an analogue life where you, you know, truck yourself down to the exercise studio.

Yeah, and I think one thing we've proven at Peloton is there is this need for people who want to work out on their own terms. I think you can draw a parallel between that and Netflix and your cable companies.

We'll draw that parallel: You want to watch on demand and you wanna get the food when you want it.

And you want to wake up at 6:30 and realize you woke up half an hour late and you still wanna do your workout and traditionally you would maybe drive to your studio but now you know your bike is right there, you can still do your work out at 6:30 and make it on time to work.

So the on-demand economy — you got all the buzzwords! You're on demand, you're cloud, you're sharing!

Yeah, yeah, metric —

So in that on-demand economy, where does it extend to from your perspective, as an engineer?

You know, I think having accessible, the hardware platform and the cloud platform, and the content being available to anyone at any moment in time, HBO Go, Netflix, on demand, you can binge watch an entire season. For Peloton, you can look at what classes your friends took, and the live class might have been at a bad time, you can take it later on. Any time you want, right, you don’t have to find a babysitter because you can just take that class right there, your kid is right next to you and that convenience is what we found attracts people to our cycling experience.

You guys talked about doing an IPO. Do you see that in the future or do you look at GoPro and just cry?

[laughing] There's GoPro and there's a few other platforms.

Bunch of them! Have had some troubles!

Yeah, definitely, I think for us we're focusing on growing and also being a focused fitness platform, not in the Fitbit style, but in you have a cycling, fitness exercise. We actually just released, beyond the ride content, which are yoga and weight training, and in the future we'll be looking at a few high-intensity interval training as well and then there might be other types of product along the same line. And we are really looking to bring this experience to a broader audience.

So as a startup you worked at other places, Cisco and Skype and other places. What is the thing that you think you have done that you have done well and what’s the thing you have done that you didn’t anticipate — the mistake you made?

Yeah, certainly, so I think what we've done well is really focus on delivering. Every single project that we work on — okay, we're gonna release the bike January 14, of course that's not the first deadline, it’s probably the third deadline, but everything we do we look at, what we can do in the short term, two or three months with a tight deadline for delivery, so that we're keeping ourselves disciplined so that we're creating features, only features that people will like. So we're creating these features, we're testing them and then we're going back and asking ourselves, is this the right feature that we created and should we create more of similar features or actually pull back on this a little bit because this is the wrong direction. And we're not wasting too much because we're evaluating on a very frequent basis, and I think the part that we didn’t do too well is um, well, I guess, Skype is not really a hardware play, this is the first time we are creating and delivering a 150 pound piece of metal. There's a lot of ...

Possible snafus!

Yes, that we did not — yes, right — you can always fix software, and the hardware cycle is something we're kind of learning on the job.

People are more tolerant of software, even if it’s awful, hence Apple Mail continues to be terrible.

[laughing] I definitely do not want Peloton to be associated with the okay-ness of that software.

Right, but I mean the hardware: People are not [tolerant]. It has to work.

Right, absolutely. So the hardware, including logistics, delivery and also all the areas of development of the hardware, I guess we are, as traditionally software people ...

But there's gonna be more and more of that combination in the real world, delivery and everything else.

Exactly, and these are areas that we've made improvements and learning continuously, but Day 1 it was a little rough, but Day some 100 and we're much better.

Excellent. Well you just let me know when we're gonna replace the human body, and we can just download ourselves and we don’t have to worry about any of this shit.

And do virtual workouts! With your virtual avatar!

Why would you work out if you didn’t have a body? Why would you eat and everything else? Well, we'll get into that later. Thank you, Yony, I really appreciate it, it was great talking to you, thank you for coming by.

Thank you so much.

This article originally appeared on

Sign up for the newsletter Today, Explained

Understand the world with a daily explainer plus the most compelling stories of the day.