TrainingPeaks CEO, Lee Gerakos, joins us to discuss the current trends in training software and where he thinks they are going.
Episode Transcript
Trevor Connor 00:00
Trevor Connor, hello and welcome to fast doc, your source for the science of endurance performance. I’m your host. Trevor Connor, here with Chris case, many of the biggest and most exciting training innovations over the last 20 years haven’t come from what we do on the bike, from from the software we use to track, monitor and plan our training. Some of us still remember when pro level tracking meant scribbling time, distance and average heart rate on a sheet of paper. Today’s training software can highlight an athlete’s strengths and weaknesses, monitor sleep and recovery, suggest workouts and provide detailed analysis that physiology teams could only dream of a few decades ago, and these platforms are only advancing faster. So in this episode, we’ll explore the biggest current trends in training software. Our guest is Lee garakos, CEO of Trainingpeaks, but this is not an episode about Trainingpeaks itself. Instead, we asked an AI tool to identify the most important current trends across platforms, whether or not Trainingpeaks uses them, and we’ll examine each what innovations they bring, what’s overhyped and where caution is needed. We’ll start with AI not to continue to debate if it’s taking over, but to consider how it should and shouldn’t be used as a tool. From there, we’ll look at integration, combining workout logs, sleep scores and biomechanical data into one platform. Next, we’ll discuss platforms offering daily assessments and recommendations, and whether they can ever match the insights of a coach. Finally, we’ll dig into gamification and the growing use of social features. Joining gracos. We’ll also hear from Dr Andy Pruitt and Ryan Ignatz, two leading bike fitters, cyclist and podcast host Jack Burke, who emphasizes the importance of mentors. Robbie Ventura, owner of the velocity training platform, who discusses data integration and pro cyclist Matt sharp, who shares how he and his coach use virtual training. We’ve wanted to have Lee on the show for a couple years now, his insights have always been really thoughtful. Whenever we talk with him, we thought it’d be great to get him on the show and have him share some of the things he shared with us, with all of you, our listeners, and thanks to our partnership now with training peaks, we’re thrilled to finally have this humble CEO join us. As I’ll mention later in this episode, when I started out tracking my data, I was literally pulling it into an excel sheet and having to line everything up just to get some decent data. I have been using training peaks or cycling peaks or W, K, O, all the different iterations and names for countless years. And one of the things I have appreciated is that their changes and improvements have been extraordinary, but they’ve also been incredibly thoughtful. I have used other software for other reasons, where they make big changes, and you just look at the changes and go, why’d you do that. I’ve never said that with their improvements, it has been iterative. They haven’t jumped on every single innovation. It’s something that they think through. It’s something that they make sure it meets their primary mission, which is to help athletes track their data and to help coaches and athletes communicate with one another, and where the innovations help that they integrate it, and when the innovations don’t help that, they talk about it. But they’re very wary that said, training peaks is adding tons of new features all the time. Even in the last year or two, they have added the strength builder, they’ve added training peaks, virtual fuel, insights, I believe they’ve added about 46 new sports subtypes. They’re allowing payment stack charts. They’ve added women’s health metrics. So it’s not like they’re not innovating. As a matter of fact, they are doing a ton. But as I said, the thing that I appreciate about training peaks is it is always very well thought out. And I think as we talk with Lee today, you’re going to see just how thoughtful he is about all this and how much consideration they give before they say, let’s add this chart, or let’s add that feature. So if you’re ready to take a deeper dive into the data and some of these new features, for a limited time, get 20% off a trainee peaks premium subscription with code fast talk, that’s code, F, A, S, T, T, a, l, k, for 20% off. So all that being said, put in your best trending hat, and let’s make it fast. Well, Lee, pleasure to have you on the show.
Lee Gerakos 04:12
Thank you, Trevor, thank you for having me. Thank you, Chris, great to be here. So this
Trevor Connor 04:16
episode has actually been in the makings for a while. There is a history to this that I need to quickly explain. You and I have done a couple winter rides together. You’ve invited me warned that you’re an age group rider, you’re not nearly as fast as me, and then we get on the bikes, and then you for three hours, absolutely destroy me. I don’t know about that tends to hurt, but when we’ve done these rides, it’s been really fascinating conversations that I’ve really appreciated. You’ve wanted to talk about the science, talk about the research. You know, I can tell that you’re thinking about all this in terms of what you want to do at training peaks, but something I’ve really appreciated is whether we’re talking about some. Thing that Trainingpeaks does or doesn’t do, you’ve been very reflective in the conversations. And even if it’s something that Trainingpeaks isn’t doing, you want to learn about it. You want to hear about it. You want to bat it around discuss the pros and cons of it. And I always really appreciated that both kind of humility and reflection that you’ve shown in all these different topics. So that is what motivated me to reach out to actually, a long time ago, to come on the show, and basically said, Just come and let’s have one of those conversations that we’ve had on these bike rides. Yeah.
Lee Gerakos 05:34
Well, thank you. I think that for me, personally, cycling has been such a huge part of my life, and not only cycling itself, but coaches. And I didn’t grow up as an athlete, really, and when I moved here to Colorado, I was just amazed by how much people enjoyed life and how much endurance sport was part of that, and the culture of it. And very early on, I just got into it, but also realized that I didn’t know exactly what I was doing, just like any journey in education, I worked with a coach, and he was fabulous. It really changed my whole experience within this sport, and that was 20 years ago, and I’m still in touch with him today, and when we got together, it was really an honor for me, because I’ve been listening and learning from you both for the last 20 years as well. And so I appreciate the kind words, but you know, right back at you, with regard to everything I have learned from you both over the years as well as during that ride, I
Trevor Connor 06:32
really appreciate you’re saying that as well. So this is going to be a bit of a unique conversation, because what’s important is we just want to talk about what is happening with the training platforms, what we see in the future. So this isn’t directly a conversation about training peaks. This wasn’t let’s bring you in and talk about what you are doing at training peaks. I know you’ll probably bring training peaks up a lot, but it’s that broader conversation about what is going on. And we’re also going to talk about things that you’re not doing at training peaks that you might be considering, but haven’t decided whether you’re going to do it or not, and it’s really just this broader conversation about there are a lot of changes happening in the training platforms and the data and how you’re seeing it, how you’re thinking about it. And like I said, sometimes training peaks will come into that and you might be doing something with it. Sometimes it’s just, hey, here’s my feeling about it, but we haven’t even considered it yet. Yeah, before we dive into all the trends and training platforms, I think it’s important to remind all of us about what we’re trying to accomplish in the end, here’s Dr Andy Pruitt and Ryan Ignatz boiling it down to something that’s really basic but really important.
Dr. Andy Pruitt 07:38
I think work is work, and whatever motivates you to do work, to me, that’s the bottom line. If it’s Zwift, I don’t care what it is, as long as you’re motivated to do work. Some people need something new to motivate them. Some of us old habits die hard. So I can’t say that I’m a fan or not a fan of the new things, because I just think whatever motivates somebody to
Trevor Connor 07:58
work hard. Interesting answer, Ryan
Ryan Ignatz 08:00
had thoughts. That’s such a brilliant answer, yeah, because I agree with the ideas of whatever motivates you to do what you need to do to accomplish that goal, I think that’s the bottom line, and I think there’s an element of mixing it up as well that often is overlooked, and people get again, stuck in a rut, or they’re just so comfortable with a certain coach often, I think they start to see a plateau. So they need something different. They need some change. So whatever that stimulus platform is, then I think, move towards it.
Trevor Connor 08:33
So the other thing I’m going to say is we are going to start with talking about the impact of AI. But this is not an episode about AI. This is an episode about all the trends that we’re seeing in training platforms and data that said AI is having an impact on everything. So I think it’s going to come up throughout this episode, not because we want to focus on AI, just because that is what is having the biggest impact right now. So I think it’s where we need to start, and I know you want to address this, so I’m going to throw the question to you, but I think it’s important to start by saying AI is a tool. I know there’s coaches that are worried about AI replacing them, and there are people out there trying to develop AI tools that can replace the coach, but there’s also people out there that are trying to use AI to enhance and help the coach. AI is not trying implicitly, to be one thing or be another thing. It is a tool that we can use any way that we want. So it can go all these different directions. So with that, I’ll throw it to you. Where do you think this is all going
Lee Gerakos 09:41
Sure. So I guess, to provide a little context in terms of where my opinion is coming from, I’ve been a software engineer and a technologist pretty much all my whole life. I started as a software engineer and really quickly found myself working in big data, in machine learning systems. I got my master’s here at CU right across. Street, and I took ml courses there, and I have seen technology continue to progress over that whole period. And each time there’s a wave, there’s always like, what is this disruption going to do? How is it going to displace everything going on? And Gartner has this, what they call the tech hype cycle, right? So there’s some trigger, technology trigger, that gets everyone really excited about it, and then you hit peak inflation of expectations, and then it starts to wane, as people start to be like, Oh, this is actually going to take probably a little bit longer than we think, to where you start to get to this plateau of, like, adoption and efficiency. And it’s incorporated. And so with AI right now, with open AI, and what came out in the last couple of years, it’s really kind of brought itself to the forefront, because we finally have something that’s very consumer facing that people can interact with and see, oh, wow. This is what this can do for us now as a tool, it’s been applied throughout even within training peaks. It’s been in the product for years, right? Like there’s a number of places in which we have machine learning today. So if you take your TSS, which is based off your power meter, that is a very discrete derived metric that is taken directly from the power meter when it comes to heart rate, tss, or run TSS, those are derived metrics where machine learning techniques were applied, right? And you see that a lot more within wkO as well, every technology, though, can be applied as a tool. With regard to AI, specifically, it’s such a broad topic, and what we need to do is also, I think, define some of those categories, so we can jump right into generative AI, where chatgpt is today, or we can talk about some of what those different categories are.
Trevor Connor 11:35
Yeah, that’s where we were going to go, is define AI, because I think it’s a term that people use. And obviously marketing has jumped on it and said, Oh, we’re using AI for this or that. And I still remember one guest on the show saying 90% of what people are claiming as AI is not actually AI.
Lee Gerakos 11:52
So at its core, what we are trying to do is create human intelligence, right? So how can we infer in reason? And if you categorize that, that can come up in a number of different ways, to your point of like, true, AI or not, what you’re probably just talking about, there are expert systems. So those are those rule based engines where a bunch of knowledge has been kind of like codified in terms of that if then else logic of in this situation, drive this outcome in this other situation, do this other thing that is separate from machine learning, which is, how do I take just a corpus of massive amounts of data and understand what is going on and identifying patterns? And then finally, you can take the machine learning, then you have deep learning, and then you have things like neural networks. And I know Alan cousins was on a couple of years ago talking about that, and that is where you allow the machine to really, kind of like, distill and pull out the features that may help derive an outcome. And then obviously we have generative AI now, which is, how can I produce new information or content based on a corpus of existing information at the end of the day, too, most of these techniques are about probabilities. So it’s not a perfect science, right? It is about what is the greatest predictor of the future? It’s past behavior. And that is something that you know, I’ve heard Andy Coggins talk about, with regard to the greatest predictor of performance is past performance. Mark Twain talked about, History doesn’t repeat itself, but it often rhymes. And so there’s kind of like this continuous theme of, how do we learn from the past and apply it to the future? And that’s a lot of what AI is trying to do.
Trevor Connor 13:22
So I was out for a ride two weeks ago with a gentleman named Will Murray, who’s on the board over at USA Triathlon, and we got into this conversation, and he made a really interesting point that I had to give a lot of thought to, which is, AI, by definition, is trying to mimic human intelligence. But what we have to remember is we’ve gotten really good at having it create an output that is written in a way that we can understand, that looks like it’s thinking the way we’re thinking, but AI does not think the way the human thinks. It thinks very differently, and then what it produces we think of as well that looks like human intelligence. And the question is, what impact is that gonna have? That we think it’s thinking like a human, but it’s not. It thinks completely differently.
Lee Gerakos 14:08
Yeah, I think what it is good for is helping us understand and this is why I think for the coaches, there’s actually tremendous benefit. Because in my experience, what I’ve often seen is that there are certain things that you just don’t think about or consider. They’re kind of buried, and they’re relationships that seem non obvious, but they actually help provide information and better understanding. I think where we can run into problems sometimes is when we talk about AI taking over. All of a sudden, it is the AI told me to do this. And there’s a funny episode from the office where Michael and Dwight are there on the way to visit a client, and they’re at the intersection of a corner, and there’s the GPS that’s telling them to turn right, right. And they’re like, No, you need to go to the corner. No, the GPS is telling me to turn right here, and they go right into the lake, right? And that’s obviously an over exaggeration of what’s. Happen. But I think that at the root of it, right, like that’s kind of what we need to understand, is that the machine is giving us information back based on its understanding. It can’t distinguish between causation and correlation.
Trevor Connor 15:12
So I had a very interesting example of this, where I started using Gemini. It was recommended to me to use it for research, to find studies. And so I was preparing for a podcast, and I asked Gemini to find me studies on a particular subject, and I came up with this list of references. I was like, Oh my God, those are amazing. And there was one study that was just spot on what I was looking for. So I went to PubMed to try to find the study, and I couldn’t find it, so I went back to Gemini and put in the title the study it had just given me, and said, Please find this study and give me the link. And then replies, this study doesn’t exist, right? And so I said, but you just gave me that reference. Why would you give me a reference that doesn’t exist? And its reply was literally, oh yeah, that was my bad. Would you like me to stop giving you fake studies in the
Lee Gerakos 16:02
future? Yeah, and the reality is, like the way generative AI works at its core is again going back to probabilities. So you ask it a series of questions, the more data you provide it, the more context it has to work from. And what it does is it says, based on this set of input, what is the response, and then what are the associated words for this kind of query and question? And so it’s not necessarily reasoning on its own. It is finding all of the relevant information to be able to provide an answer back. And it’s great information. And that’s where I think you, as the expert, really are able to discern that this is something real, and this is something that is not, and that’s where, I think, where we are now, and it’ll continue to get better, right? Like, there’s no question about it, that it is going to continue to provide those sources, to be more clear, to be more understanding, and to provide better information. It’s not to say that the information that it provides, in many cases, is really great today, but it’s still a tool that I think is really for the expert. You know, as an example in radiology, machine learning has been used for years in AI and it’s to help the radiologist diagnose the X rays. But you don’t go into your hospital today and be handed an AI printout of like, what the information is and say, Go on with your day, right? Like you still need that human to be able to understand the contents of that information and say, Yes, this is good, or no, it made a mistake right at the end of the day. We’re all imperfect and unpredictable, and so we can’t expect the information that is coming from a machine to also be perfect. But on the flip side, there’s plenty of cases where I’ve used Gen AI to really find interesting information to think about. Like yesterday, I was looking at what is the optimal placing plan to go Flagstaff, and it provided an answer back with regard to how to pace the initial part of the climb, the mid section and then the latter sections against that. And then I asked it some questions, like, why is it the optimal pacing? And it provided a rationale with regard to why that would be. And it said, Here’s also some different ways that you can pace it. And so here’s a steady state pace the whole way up. And then also, here’s what most people do, which is overshoot the beginning, when you’re going up the steep section in the beginning, past Gregory Canyon, and then you kind of fade through the rest of it. And then it provided an approximation of what it thought those different durations might be based on a person of a certain weight and a certain threshold. And then also some like physics characteristics around like, CEA and what have you. So it’s really interesting, right? Because it’s like, that’s an input that I haven’t gone and verified it though, to say, like, oh yeah, this does work, but it does provide some, like, interesting things to think about that I may not have thought of otherwise. And that, to
Chris Case 18:34
me, is one of the interesting things about the responses that you get can have mistakes, inaccuracies and so forth. But if you are not expert enough to spot them, then you just accept what you’re given. Your instance, Trevor is not the greatest example, because you went and said, I want to find this. I want to read the actual study. I don’t want just the title, but say you were, oh, I got to put together a bibliography for this, because I need to turn this term paper in, and I need to show all my work. And you’re like, hey, generate a bibliography for me, and it creates a bunch of fictional references. It sounds great, but until you verify that you don’t actually know that it’s right. You know one
Lee Gerakos 19:19
thing that I think especially amongst this group, we sometimes take for granted is how much we know at this point. And I think that is where a new person coming into training, into cycling, into endurance sport, it’s an overwhelming experience to begin with, and at this point, we’re also, like overloaded with so much information from so many sources and so many differing opinions that it’s then hard to distill it in terms of, like, how do I move forward? It’s almost paralyzing at that point too. So I think that an important piece here is, what is it that we’re trying to understand? What is it that we’re getting out of it? What are our goals and from? Where are we coming to it from? Right? So I think as a group of people. Who geek out on this stuff a lot. It’s super fascinating to interact with. It’s super fascinating to see what it comes back with. But really, what it’s doing, in a lot of ways, is spurring ideas.
Trevor Connor 20:10
So that kind of brings us back to how this is involved with the training software. And I’ve read a very interesting article getting ready for this that kind of said that, which is, coaches and athletes have gotten very data obsessed now, and the problem we have is there’s so much data, it is very hard for a coach to process that data. That’s not what we’re built for. Same thing for an athlete, true. You know, it’s can get tiring going through 300 different graphs of different things and trying to say, what do they all mean? I think that’s where the AI can be very helpful. But what it lacks that we’re talking about here is judgment. It can give you interpretation, it can process all the data, it can find the trends, but it doesn’t have that judgment to say, what does this mean? And have the experience to say, yeah, no, I get what’s going on with this athlete, or I get what’s going on with you as an individual, and that’s what we still need to do, yeah,
Lee Gerakos 21:04
and that’s not unique to coaching and endurance sport. I mean, that’s really true across all domains and disciplines. As I mentioned, I’ve been working in big data machine learning for years, and the reality is that it can do a really good job at predicting things. The incremental performance, in some cases, is marginal, but it can mean a lot depending on the context right, like at the World Tour level, a 1% difference is massive. However, at the age group level, there’s probably a lot of things that I can be doing first to get my maximal gains before I’m worried about those things that are really at the elite level.
Trevor Connor 21:34
So that really brings us to one of the questions I ask you is, How good can the AI be prediction? I’ve had this conversation with you. I’ve had this conversation with many people that right now we’re in that phase of interpretation and analysis, but the ultimate goal is prediction, to be able to look at all the data and say, here’s where we think the athlete will be in a month. If we throw this particular type of workout at this athlete that’s going to get them on top form for their race. It’s predicting the future. How good do you think AI is at taking that data and being predictive?
Lee Gerakos 22:11
So that’s an interesting question, because in a lot of the reading research discussions I have had as well, when you try to apply things at a group level, there’s so much variation in terms of an individual response, and how do you then cluster and cohort those groups of people relative to one another, that I think those predictions are plausible, but they’re going to again, be imperfect, right? So how much can you increase someone’s threshold, power up a climb within a given period of time? I think there are certainly models and predictions that can inform that the area where you have to be careful is that, again, it’s not going to be 100% accurate, it’s going to be directional and provide a target for someone and a number of things have to go to plan to be able to have that play out. The question that I’m not sure of is, how much more lift do you get in terms of that AI prediction versus what we have seen with regard to some traditional methods and existing metrics that are in place today to compare those two things, and I’m not aware of any studies that have really gone into that to say, like, what is the comparison between an AI driven like, at that top level of performance prediction relative to the coach prescribed one. Now there’s a an argument to be made there of like, combining the two together again, which is where that expert and that data continues to help inform and it’s, how do I get to a quicker decision making? Right? Like, at the end of the day, it’s no different than saying what I want to know is how to multiply these two numbers together, and what the value is not to do the math myself on a sheet of paper in my head, right? Like I want to get to the answers very quickly, and that’s where technology is certainly going to continue to help and accelerate us.
Trevor Connor 23:53
So I’m going to throw this at you as a bit of a challenge at you. I wanted to give AI actually its own opportunity to get into this conversation and potentially defend itself. So I did go to Gemini and asked it the question about, what is the impact of AI on training platform software, and it wrote a very long report to me that I should send you, because it’s very interesting. It was written by AI, so take it with a grain of salt. But first of all, it said training platforms, or training software, is really being defined by two platforms, training peaks and Strava. So there was a compliment to you, but then it got a little mean towards you. And what it said is, your platform is based on concepts of banister and Coggin. So banister impulse response, which is basically the basis of the PMC, and it said, actually, there’s no research basis to show that those work and are effective. And said it’s static, and the training platforms are a little too static, and AI is going to allow us to have a much more dynamic experience for the athlete where. Can modify it day to day, which I found interest an interesting flip, because that’s always been my criticism. You know, where coaches are essential is coaches are the ones that can be dynamic. Now, here’s the AI going, No, no, no, coaches and the training platforms are static. You’re only going to get dynamic if you use us so certainly being very strong and defending itself.
Lee Gerakos 25:20
Yeah, I think when it comes to that, I think that there’s always nuance. This is not an either or conversation. It’s not a zero sum game. And I think that sometimes we talk about these things in being a very binary decision of, do I use this? Do I not use this? And so from my perspective, it’s like it adds value. The question is, though, and sometimes, is that adaptation? That adaptive plan being an age group athlete who has worked with a coach and worked on my own, when do you actually need to continue to push through? When should you be doing it, regardless of how you feel? Does it have the context and the information to be able to do that accurately? Does it know you and what you’re capable of isn’t intentionally trying to push you because it wants to see what those limits are, right? Like, one of the things that has always been so fascinating about training, why I think so many people gravitate towards it, is, number one, you feel alive, right? And part of that human experience is, what am I capable of? And you have different personalities, and different people go through this differently. And some you have are that, like, I actually just go too hard all the time. And then there’s other folks that you know, they’re not sure what they’re capable of, and so they might hedge towards being more conservative. And then you have folks in the middle right, like you have that broad spectrum, and I’m not saying any of them are more right or wrong right, like there’s no judgment in that. I think it’s just like we all have our own human experience. And I think what’s challenging sometimes is to rely on data coming in to just tell me how I’m supposed to feel. One of the things that I have actually always loved about training, why I keep going at this is it’s just helped me understand myself more and more. And so if AI can help do that, if that’s part of the conversation within there, and I’m not second guessing myself around that every single time like a change occurs, I think that’s awesome. I think that you have to have correct data going in to be able to make those decisions appropriately. You have to have a tremendous amount of self awareness to be able to make those decisions, and at the end of the day too. Like every coach I’ve spoken with, every expert I hear from at the end of the day, it still comes down to like, how do you feel right? Like, how are you thinking about this right now? How motivated are you the coaches that I speak with, it’s interesting. You know, they’ll collect a lot of these metrics, but they still use that acute training load and that chronic training load to be able to be able to understand those ramp rates and the decision making within that and it helps provide context against everything. But it also comes down to the athlete and how they’re feeling. I can tell you that you know, even personally working with a coach, we didn’t do that much testing because it was used as a guiding post, and if I overshot my intervals, because I was feeling great that day, but I still completed them and they were consistent, that was what mattered. So you can kind of know those differences, but it can be very confusing. And if a plan is constantly changing on you, and you’re not understanding exactly why and what it means, you can shortchange yourself, and you’re potentially like, you know, not even getting the opportunity to learn and see what you’re capable of. And so there’s that balance there.
Trevor Connor 28:25
Yeah, well, that’s gonna bring that up, because I think a lot of us see the ultimate goal here is to have that extremely dynamic plan where, literally every single day, we’re deciding what do you do that day based on how you’re feeling, how you’ve been responding to training and everything else. But I’ll actually back up a little bit and say there is a real value to having a long term plan that the athlete is aware of, and making sure that while you’re adjusting day to day, if they’re feeling off whatever you make those adjustments, they’re also sticking to a plan, as opposed to every day not knowing what they’re going to do the next day, until that day comes around.
Chris Case 29:03
I kind of wanna go back to when you’re talking about the response that you received from Gemini. My question is, what’s the source? Where is it pulling that information from? Is it inspecting how training peaks actually works and making a judgment on how it works, or is it reading the internet, reading reviews of that product, and saying it’s static, because DC Rainmaker said at some point it’s static and blah, blah, blah, so then you respondedly about how it’s about the data that goes In, and how you know that has to be good for AI to spit out good data in a simplified sense. So my question is, if you’re integrating AI technology into training peaks, what is the source of information that it’s using about a. Particular athlete, is it their data? Is it trends that you’re seeing in all of the data that’s ever been uploaded to training, peaks? How are you making sure that the data that it’s using to predict what that athlete should do next is quote, unquote, as clean as possible?
Lee Gerakos 30:17
So I think at the end of the day, we need to be thinking about motivations, right? Like, AI is very popular right now, and so you see this race of hype around, like, how can I sell this? There’s so many technology companies invested here. There’s so much opportunity invested here. We saw that with the Internet. I saw a video pop up the other day, and it was the 1983 video game crash that occurred, where there was so much hype around video games, and it was flooded in the market, because everyone saw this, like, cash grab. And then obviously, like, video games are huge industry today, so there’s tremendous value there. And I feel like that is some of where we are today. I think when you think about training peaks too, as a product, it’s for the people that love to really like dig into this stuff and to learn and to understand what is the science of training, and so we need to be thoughtful around that, because I don’t want to take that away. You know, at the end of the day, we’re also, like, 100% supportive of coaching and the human coach, the human relationship, and everything that comes along with it. It has changed my life. I see it change countless lives on a daily basis. It makes it a wonderful experience. What I don’t love is this conversation of it’s either AI or it’s a human coach, versus, how do you leverage AI to improve the coaching experience, to be able to focus on the things that matter more and then the things that don’t when it comes to some of the modeling, I would say that a lot of the high value features that are AI driven exist in wkO today, and a lot of those features we are talking about bringing them into training peaks. So your time to exhaustion, your VO two Max, your modeled FTP, your FRC, right? Like, how can those things help again, further inform that training, and that’s also a big part of what makes it special. And so for me, I want to ensure that we preserve that. And so what we will always do is help inform. We will help to understand. But the essence of what is that coach, athlete relationship, what it means to get expert instructions. There are so many coaches out there that have podcasts, that have content that share their information. We want to continue to support that, because that’s what makes this whole industry better. And so to your point on, like, where is that source of data coming from? It’s coming from all those articles that have been written over the years by all of those coaches that are out there. And so we want to make sure that we continue to support that as an ecosystem, as a culture, as an industry. I mean, you guys correct me, but I don’t know of an athlete, especially at the top level, that has succeeded without the support of a team and a coach, so I don’t have an exact answer for you, but like, the last thing I want to do, and the last thing I think is a viable thing to do, the last thing that I think is good for anyone is to say, how do we just take this data and just apply it everywhere, like I think you’re just not going to create an experience for anyone that is going to stand the test of time, that is going to be good for the culture of sport and good for people.
Trevor Connor 33:10
This is a good place to hear from pro rider Jack Burke, who talks about the importance of the human connection in training for
Jack Burke 33:18
people that, let’s say, if they’re using a training software and they’re mostly relying on that, if you don’t have access to hire a coach, like, if you just can’t afford it, or something like that, or you don’t have access to a good coach, I would say, instead, shift to building a good team of mentors. Like, reach out to people who have been where you want to go. Like, for me, it was I just accepted when they talk, I shut up and listen, because, like, they’ve been where I want to go, and I listen to these people. And I think that’s important, because if you’re coaching yourself, and you’re using all these tools, yes, you can have all this great data, but you will have some blind spots as far as things you’re not thinking of, and it’s good to have that external perspective. Of course, having a coach is the best option, but if you don’t have access to a coach, a really good skill is learning how to reach out to mentors and learning how to ask intelligent questions and be respectful of their time. And be respectful. Means short, concise, specific question, not vague, like ask them something you can’t google learning how to ask good questions. You will get the best information from some of the best riders and coaches in the world that way, if you know how to ask, well, that’s probably, let’s call it like the best, like, sort of free solution if you’re just relying on training softwares and stuff. But I mean, you can do a really good job SELF coaching yourself, but I do think no matter how good these tools get, you do need that human connection, and especially when you’re starting, because when you’re just started, like learning how to understand this stuff, you need someone to teach you how to use the tools. When you get later in your career, and you’ve been coached for a while, you sort of know a lot of things, it’s still really valuable to have a coach. I mean, for me, it was just to keep me from overdoing it, or just like it gave me so much bandwidth that I didn’t have to think about my training. That was one of the best things about working with Steve. I never thought about my training until I was putting my shoes on. I just look at whatever he gave me, and I just do that. Versus in the past, I’d be constantly thinking about it, and I always got to do more, and I’d be focusing on the wrong things, and I wouldn’t focus on my weaknesses as much because I just didn’t
Trevor Connor 34:57
realize them. I love that. You brought up the video game cross. Ash in 1983 because a couple weeks ago, it was out for a ride. The podcast I normally listen to had gotten through all of them in the middle of a climb, and this podcast comes on about the ET Atari video game. And I’m like, This is stupid. As soon as I get to the top of the climb, I’m gonna switch by the top of the climb. I was fascinated, because apparently this game caused the entire crash in 83 because they put so much money into it and was such a big flop. Yep.
Lee Gerakos 35:30
It was very rushed. If I’m not mistaken, I believe there is a landfill in New Mexico. I think there’s a Netflix show about them going to find that, right? Yep, so with
Trevor Connor 35:40
the ET game. So, yeah, never knew about any of that.
Lee Gerakos 35:43
I lived it firsthand. What’s the relationship to the landfill? They had so many games that they just never sold and they married them all.
Trevor Connor 35:52
The short version is our sahari had this game developer who was apparently amazing, had developed these huge games, and then Steven Spielberg came to them and said, can you develop an ET game, and they need it in, like, six months? And typical game development then was like a year or two, and this guy’s like, yeah, I can do it, and I want no help. I can do it all by myself. And he basically locked himself in his house or whatever, and just worked 12 hours a day to come up with this game, and it was not a great game. Like I said, it’s not the worst game ever, but it was not a great game, but it got really hyped, and every kid got it for Christmas. And we’re just kind of, yeah, there was a lull in the home video game system because of that. Yeah, it’s
Lee Gerakos 36:37
interesting. So Nintendo, right? Like when they brought back out the first Nintendo Entertainment System. That’s where it started to come back. And when they designed that for the American market, one of the things they did they had a top loading cartridge tray that pushed down because they wanted it to look more like a VCR, as opposed to a video game system, you know, pushing cartridge. And they also sold it with Robbie the Robot because they wanted it to be more of a learning system. And so they really took this conservative approach of like, how do we show the value of this? But then we all bought Super Mario Brothers and it was all history.
Jared Berg 37:09
Healthy eating can be a challenge. Sports Nutrition is all about knowing what to eat, when, how much, and avoiding temptations that can derail your plans. I’m Jared Berg, and I’m a registered dietitian and exercise physiologist with fast talk labs, I’ve created a new eight week sports nutrition course to help you break old habits and establish healthier high performance feeling practices for years to come learn more and get started on fasttalklabs.com look for athlete services and sports nutrition.
Trevor Connor 37:39
Going back to our conversation. I was really glad that you guys just had that discussion, because the last thing I’ll bring up here after that, AI response absolutely attacked training peaks and Strava. It then actually showed some humility, and it got into that predictive ability and said, The one issue here, which I agree with 100% is there’s very little research on the long term effects of training, and there’s a weakness that most of the research has been done in the lab, not out in the field, and out in the field can be very different from the lab. And so it said itself, AI’s ability to predict is limited because we don’t have that data, and we need to rely on experience, Coach experience. So it even said
Lee Gerakos 38:24
that, yeah. And to make it fun, to your point on the lab like that actually reminded me of you’ve shared this a couple of times, and one of them that I looked up was you were in Ithaca, right? And believe your coach was a physiologist, yes. And there were two stories. One was you had this, like, two hour or four hour ride to do, and it was like snowing outside, and you just said, I’m gonna go outside that much, and then also, like, you’ll have to fill in the gaps here. But I seem to recall something about it was like very disciplined workouts, where it was like robotic in nature, in terms of what was prescribed.
Trevor Connor 38:56
So this is where I went. I think I need to change coaches, yeah, because we had a weekly time trial, and a guy who’s won the national Time Trial championships twice was showing up, Glenn Swan every week. And so I had an opportunity to go and raise somebody that strong. And so as a coach myself, I would go, if anybody said I’m trying to build for time trials and we have this on a Thursday night, I’d be absolutely we are building that into your plan, go and do that. My coach didn’t want me doing that because there were hills on the course and that would mess up the data. So he instead wanted me on Thursday nights to go and ride this road that was about 10 minutes. So you want me to go out and back and just keep doing laps on it, because it was a perfectly flat road, so he’d get good data, yeah, and that’s where you’re to me kind of missing the point, right?
Lee Gerakos 39:47
And honestly, like, this is something that I think we’ve seen in the evolution of coaching and training, frankly, where looking at some of my old plans and working with a coach, a lot of it was especially. You once you got to the summer, here’s this group ride. Here’s how I want you to ride this group ride. I want you to sit in. I want you to attack off the front, like that incorporation of fun and experience and also knowing that, like some of the things that I really needed to work on were my pack dynamics and my confidence, frankly, right? Like it was, did a lot of the work now take these opportunities to really express all of the work that you’ve done in these fun environments. And I think that sometimes when we talk about structured training, specifically, we have kind of conflated structured workouts with steps on them that are very rigid with structured training, which is having a goal and an intention and a purpose to drive towards an outcome. They’re different.
Chris Case 40:43
How about we shift gears a little bit, and let’s look into the future. Use your crystal ball, if you will. You probably have a better sense of where your company is wanting to go and going. But let’s talk more broadly about data integration, for instance, and what that looks like in the future.
Lee Gerakos 41:00
Sure, sure, there’s always new sensors and devices. I think one of the things that has really been at the core of training peaks over the years, and you see it within the industry, is that it’s an open ecosystem. So there’s two parts to this. There’s always new sensors and devices being developed. And so part of it is, how do we ensure that we can get that into those experts hands to understand the relationship of that data relative to other data that is well understood today. So Dr Steven Seiler, who’s been on your show a number of times, he’s working with time where, right, I’ve heard you all mention this, and so that is a strap that goes around your chest to understand respiration rate and the effects of that within your training. So that’s a very interesting potential for that device. I believe it’s even used by a visma Lisa bike right now, at least we’ve seen it within the races. And how does that help inform training and so on, that open ecosystem and what is always being brought in? One of the things that we’ve really worked on in this last year is to say, how do we ensure that as these new devices are made available. Typically, the first place in which that device will be made available is within the head unit of the device, recorded within the Fit file and then brought into the platform. So what we wanted to do was make sure that it was a no op effort to continually bring that data in where previously we were stuck with static fields. Now, once that data is in there, we can start to understand the relationship of it. We can have those conversations with those coaches in terms of, like, this is how we see being able to apply these metrics, right? Because, to your point earlier, what we want to always do is make sure that we are listening and learning and understanding of what’s out there, not necessarily dictating and so like, that’s also one of the challenges with training piece, which is both kind of like its strength and why it gets criticized so much, is because we want to be agnostic. To a large extent. We want to be that place where training and science can continue to develop and we can continue to learn on what adds value and what doesn’t. Then there’s the other side of it. So there’s a lot of recovery devices and wellness metrics, right, like that are being utilized today. And it’s interesting, because, again, I think that nothing has been that silver bullet that unlocks a new era of training. What it has done is helped you inform and call out things. So taking HRV as an example, a lot of cases where I’ve heard from coaches that it’s been really helpful is to help form habits actually, right? Like, what is the effect of alcohol on your recovery and within your training? And now provides an objective metric through which you can see that on a day to day basis. I’ve heard from a number of coaches that say, Yeah, it’s interesting, but I’m not necessarily going to have it dictate what’s going to occur today.
Trevor Connor 43:47
So we had MOLLY BREWER on the show earlier this year, and she had a really interesting study where they did a lot of interviews with coaches on data analysis. This is coaches in all fields, and something I really remember from that study is you had coaches that were taking, you know, they were seeing the value in the some of these new devices, but they were taking the data from these devices, and in this day and age, they were still pulling that data back into an excel sheet to do the analysis. And there were two reasons why. One was often these tools came with their own proprietary software, and that software didn’t show the coaches what they wanted to see, so they wanted to pull into Excel so they could do their own analysis. But the other thing was, because it was all proprietary software, they couldn’t take the data from this device, the data from that device, and compare them, so they had to pull it into an excel sheet to do the comparison. So I’m interested in your response to this, because on the one hand, she said the coaches are saying we want one platform where we can see all this data together. But at the same time, how do you address that challenge in the future of allowing the coaches to see the data the way they want to see it?
Lee Gerakos 44:58
Yeah, I think that what you’re talking about. About is having things curated in a way that is easily reviewed, processed, digested on one side and then on the other side, is ultimate flexibility of bringing together myriads of information that continue to evolve and adapt over time. So again, like we look to strike the balance to some extent, but we’re not going to be perfect within that certainly, there are a number of cases where we do have an API that is leveraged by certain organizations to be able to pull that data together for their needs.
Trevor Connor 45:30
And is this something where perhaps a platform like wkO allows a lot more of that flexibility to look at the data the way you want to see it? Yeah. I mean,
Lee Gerakos 45:39
I would say it does rightly, because you have the ability to create expressions as metrics come in, you can build your own charts and again, like, what makes it so powerful is also what makes it a somewhat high barrier of entry for being able to use it right like I watched your wkO video, which is fantastic by the way around. Like, what are the charts that matter? Being able to share them, being able to, again, build that community of science and culture around that as an analysis tool. But it’s not the end all be, all of like, every possible metric that may come into existence over time. So again, it’s so broad that it’s hard to say, like, here’s the one solution to that, and I think that is a key point that we didn’t touch on earlier. Is this space is continually, rapidly changing, whether it’s the data itself, whether it is the tools and the technology, whether it is, you know, the AI, how do we make sure that we don’t engineer ourselves into a corner such that it’s really great in the moment and then we’re stuck, because that’s what you see in a lot of technology companies over time. And I’ve seen that, and I’ve had to unwind it, and I’ve been a part of it, and it’s just it takes a very intentional approach of saying, like, how are we developing this system to solve for those customer needs in a way that doesn’t also, like, freeze us in time?
Trevor Connor 46:51
So there’s certainly capability, but connecting it with something like, W, K, O, you know, right now, you still have to code it yourself. Yeah. So if we get to this point where all the data is being pulled into one platform, are there things that we’re gonna be able to do that we can’t do now? And also, are there dangers with that?
Lee Gerakos 47:09
Very likely, the more data you bring together, the more potential there is, right? So, and I think we need to separate potential from practical application. And even with more data like, more is not better, and I think that a lot of times we can create just more confusion in the process and more misunderstanding. Again, like even going back to what is available within a chatgpt and all the sources that it’s pulling from, there are things that conflict all the time, and there’s also people that are very adamant about a position. And so it’s very challenging to say that, like, this is the right thing to do, even with HRV, right? Like, different ways in which it’s collected, right? Like, there’s some very strong opinions around, like, the exact mode and time and approach towards collecting that data. So now you collect that data, and it’s all been captured in different ways, and now you need to make sense of it. And one of the interesting things in working within the data space is that they talk about like 70% of the time plus is spent on data wrangling. So it’s not about the modeling, it’s not about the AI, it’s not about the machine learning. It’s about ensuring you have clean data that doesn’t have bias within it, that is representative and that provides a good sample size to be able to ascertain information from
Trevor Connor 48:21
here’s Robbie Ventura why he feels there’s enormous benefits to integrating the data.
Robbie Ventura 48:28
I think training peaks is at the cutting edge of that. Obviously there’s a lot of great stuff out there in terms of analyzing power output. I think the future is analyzing all these other metrics that we have now, weight, heart rate, HRV, temperature, respiration rate, adding all of those into the mix to help us really understand the cost of a given workout or the recovery the next day or the day after that, those elements become super important as it relates to how much load someone can manage. I think that the recovery metrics, the sleep metrics, are really gonna impact training plan development in the future. And I also think there’s gotta be, to me, the metrics in ride also are gonna evolve, like the velocity platform we have in ride metrics that help people understand a little bit more about what’s happening to them as they’re actually riding. And I think to help quantify aerodynamics, for example, while you’re riding, quantify movement while you’re riding, those elements will also help guide us and allow us to use technology to better coach the athlete.
Trevor Connor 49:35
So again, going back to Molly’s paper, I’ll raise another issue that was quite surprising but interesting that paper is that coaches said, when you bring all this data together, there are privacy issues. I can now tell when my athletes have gone out drinking, I can tell all these things about their private life that I shouldn’t know from the data. And that’s the other question. And bring all this data together, do you start getting into privacy concerns?
Lee Gerakos 49:57
Well, there’s inferences that are being made. So. Obviously like, if a coach is communicating that, or an athlete is communicating with that, with their coach, they know those things. I think what is important is that there’s always consent, and so when an athlete is establishing that relationship, they are forging an agreement with one another around that data and that system and that sharing of information, right? So an athlete has the ability to not provide that information, because a lot of those things are not by default, brought together, but it depends on again, are we talking about the age group, or are we talking about where it’s someone’s job? And those considerations actually can vary greatly.
Trevor Connor 50:33
So another big trend that we are seeing in the training platforms and also with devices. So for example, I have the newest Garmin watch, and it now gives me a morning and evening report. Is this assessment, this whole idea that you can now go to the platform and go to your watch or go to whatever in the morning, and it will say, Here’s how recovered you are, here’s how you slept, here’s where you’re at with your training status. And then often give you recommendations. So for example, my watch this morning said you’ve really been beating yourself up your training readiness. It gave me a score of 10 out of 100 it said you need to rest today. Okay? And admittedly, I got to the office this morning I went, Yeah, it was right.
Lee Gerakos 51:16
What is your experience with that, generally speaking?
Trevor Connor 51:20
So this is the conversation I want to have. And what I will say is, if you had asked me that question six years ago, I would say, I take it with a giant grain of salt. I would get these assessments and just go, No, I get why it came up with that conclusion. But it’s just not right, and I’m not going to change my training based on it, in the last year, I would say the assessments I’m getting, I have been shocked how often I initially wake up, look at the assessment, go, I don’t agree, and then two hours later, going, yeah, the watch was right, so that I think they are getting better, but I’m still not at the point where I’m saying I’m going to trust it over my own judgment. And that’s kind of the question I wanted to throw to you, is, can we get to that point? And is there a point where these tools could actually do a better job than a coach or the athlete, and saying, here’s where you’re at, that
Lee Gerakos 52:15
is an unknown in the sense of, I can’t predict how well someone knows someone else and what they know about them. I can’t predict how someone knows themselves, and I think there’s a concern I would have like I think it’s always great to take an input and to help use it as a tool to self reflect, but I think there’s also danger in relying on someone else at the end of the day to tell you how you feel. Someone can pat you on the back, someone can provide guidance. Someone can provide that input and say, you know, you really need to take a rest, right? Like, there’s obviously extremes within that, but I think it’s still so important for an athlete also to use that input and say, yeah, like you just said you’re right, you know what? I’m actually going to just test it out here today and see how I feel. And I think you also touched on a very important point here, is that these things have been around for a while, and they are continuing to get better, and they will continue to get better. I mean, that that is just progression of society, right? Like everything that we have around us, we do not live the same existence that we did 10 years ago, 20 years ago, 50 years ago, but we’ve continued to adapt. And I think what’s important within that too, is that there are still, like, certain fundamentals that continue to exist and will always continue to exist. You know, side thing I thought was pretty interesting was on, I think I was watching CBS Sunday Morning a few weeks back, and they were talking about the resurgence of malls, right? Like people are going back to malls again, and we had to go for back to school, and I was surprised to go in, and it was like packed, because people want to be together. And when you talk about motives and direction, and what is someone’s angle of pushing something forward, think about during the covid era, what you heard from, especially from the investment community, from the technology community, was we have jumped ahead 10 years, five to 10 years with regard to how society is going to be. We are all remote first, like we have no need to go outside anymore and connect with one another like this is a permanent change that has just been brought forward, and three of us are sitting here in the same room now, because the conversation is just that much better.
Trevor Connor 54:18
Yes, I feel like I’m a little ahead of people on this, because I work from home for so many years as a coach, and I still remember in think it was 2016 I literally, one morning just said, I can’t do this anymore. I walked half a mile down the road to a Regis office, walked in and, like, rent me a desk, and it was worth every penny. Yeah, that being somewhere else, being around people. It’s great working from home for a few years, but there is a point where you go, I can’t do this anymore. I need to be around people.
Lee Gerakos 54:49
Oh, absolutely. One of the first events after the pandemic, I did was the firecracker in 2021 and they have that parade that goes out of town before you start the race. And I. Can just tell you, like, the emotion that everyone had, there was so much gratitude that we’re all together again. We could experience things together, like, there’s just nothing better. And so, you know, I know we’re talking a lot about technology and a lot of the advancements, but like, there’s still, like, these fundamentals that I think are so important to not forget about, and not just say, like, oh, we have technology now we no longer need these other things. I think again, it’s not an either or conversation.
Trevor Connor 55:25
So keeping that in mind, what would you like to see in the future with these assessments? How do they best balance that human side with giving you information that you might not be able to glean about yourself?
Lee Gerakos 55:39
Well, I think it’s that combination of subjective and objective information brought together. And then also, you know, this is another area where AI can and will help. This is something that we’ve talked about internally, is, I think grant talked about it the other week when I was listening to the show around, like the mood of someone, right, like it comes through in the contents in those conversations. So you have your objective metrics with regard to your HRV and your sleep, and then you’re seeing someone being short with you, right? Like those two things together are very strong signal there’s something going on. Now, does that mean that you should stop doing what you’re doing? I don’t know. It’s probably an opening of a door to a conversation of what’s going on, right? Because, again, like in that coach athlete relationship, and I think that this gets lost sometimes. It’s just like anyone that works together, right? Like you guys are colleagues, but you care for one another, right? Like people care for one another, and coaches care for their athletes tremendously. And I think that gets lost sometimes. And so yes, there’s a piece of it where it’s about performance and about improvement, but there’s a genuine caring for one another in terms of what’s going on, and I can think back to certain things that I was going through, and a coach seen it kind of show up, right? Like what’s going on. You’re not thinking like a winner. You seem stressed. And being able to have someone objectively look at that and wake me up a little bit. It was so valuable. And I think there’s so much to that.
Chris Case 56:59
Fantastic Trevor, can I put you on the spot and ask you, as a user of many devices, you’re kind of a tech guy. You like to try stuff out. You give it the full shot. You give it some time. You know that they evolve over time and progress over time. What would you like to see further integrated into Lee’s product that would help you make decisions about whether you’re ready to go or not. I think
Trevor Connor 57:25
it’s similar to what you are saying, which is, right now, it’s just taking a bunch of data then saying, here’s where we think you’re at. I think ultimately it should be a conversation with the athlete, whether the AI can do this or not, I don’t know, but where it can say, here’s what I’m seeing in the data, and it can ask you questions, how are you feeling? Where do you think you’re at? And then ultimately generate the report based on both your qualitative assessment along with the data that it brings. So it’s a back and forth where it can say, well, you’re saying you’re feeling pretty good, but here’s what I’m seeing in the data that’s different. And it can point out when there’s contradiction, it can point out when the data is backing what you’re saying. And together, you can come up with that assessment. Because I don’t care how good these tools get, there will never be a point where just based on data, it’s going to get the whole picture. And there has to be that athlete assessment. How do you feel?
Lee Gerakos 58:23
Yeah, I think, again, going back to probabilities and outcomes, one of the things I think that makes human relationships so wonderful is that there can be conflict. Not everything is perfect all the time, and I think that’s one of the things that we’ve seen with AI chat bots even today, right? Like you can, quote, unquote, now have the perfect relationship, and I don’t know that that necessarily helps us right, because I think that, again, what is so important for us as people is that things are going to go wrong. And do you have that person or people, that support system that can help you work through these things, that can help you show the things that you might be missing, the non obvious things. And so the only way that that can be possible with AI is if it literally collects every single thing about you and all your thoughts, and then all of a sudden, we’re like in 1984 George Orwell, right, yeah.
Trevor Connor 59:17
Well, I’ll give you a bit of an example of all this that kind of pokes fun at me. We did that episode with grant that you mentioned, we were talking about irritability, and you know, I gave the example of my girlfriend pointing out that I was being a grouch. So I’ll give you another example. I was hanging out with her yesterday, and she started singing this song to me about how she likes it when I’m not a grouch. And I was a grouch on Saturday, and she likes me much better when I’m not one, which is why I woke up this morning thinking, Oh, I’m great. And then, you know, my watch says, No, you’re not recovered at all. And then I remember her singing that song to me. I went, nah, okay,
Lee Gerakos 59:52
that’s right. And I would say, I’m curious, from your perspective, if you heard that from a machine that said you’re being a grouch, i. How would you have felt about that, relative to your girlfriend telling you were being a
Trevor Connor 1:00:03
grouch? Can I sing as well as she can?
Lee Gerakos 1:00:07
What I’m saying is that, like, did it mean more to you coming from her? Yes, right, absolutely. That’s what I’m getting at. And I think that’s one of those things where you can’t deny that hearing something from a computer, I can just ignore it.
Chris Case 1:00:20
I think that’s true in 2025 Yeah. I think the sad part, maybe, or the just the natural progression, is that we’re all going to get more and more used to hearing feedback from machines that we actually say, Oh, that’s a valid point, good point machine, Mr. Machine, or Mrs. Machine, or whatever, and that will have some impact on us. Now, it’s a little easier to dismiss, and we can say our girlfriend is more important to us their word
Trevor Connor 1:00:51
machine, but maybe not always. I fear the day where she starts singing to me about being a grouch. I’m like, No, you’re wrong. The AI told me yesterday,
Chris Case 1:00:59
yeah, exactly, yeah. When the roles reverse
Trevor Connor 1:01:03
dismissed, I don’t ever see that going well.
Chris Case 1:01:06
Not well, not well.
Trevor Connor 1:01:09
Built for coaches from day one, training peaks is the complete coaching solution for 25 years, training peaks has evolved with endurance sports, making the planning and analysis tools you love even better year after year. And it doesn’t stop there. Training peaks has added strength builder. Training peaks, virtual fueling, insights, 46 new sports subtypes, payments, stack charts, Women’s Health Metrics and more. Ready to take your data deeper for a limited time. Get 20% off a training peaks premium subscription with code, fast talk, that’s code, all caps, F, A, S, T, T, a, l, k, for 20% off. So the last two trends that we’ve seen in the training platforms that I think we need to at least touch on are one this trend towards gamification and augmented reality, and I know you’ve now gotten into that game, and then the other one is bringing a lot of the social media features into the platforms, and Strava is obviously the best example of that, where, in some ways, it’s I remember my brother saying, yes, Strava is my social media platform. So what’s your feeling on the future of both of these? Let’s start with the gamification and where you feel that is going so
Lee Gerakos 1:02:21
it’s interesting because, you know, I was doing a bit of reading on this, and just the notion of gamification, right? Like, there’s a lot of areas where gamification can be very helpful. It is used in therapy. There’s opportunities to turn something from being just pure work into something that is more enjoyable, I think, where you have to be careful about it, and we all have obligations, especially for me, right? Like as responsible on a platform is, when is gamification, a reward system that is positive, and when does it become manipulation? And I think that, like we all have a responsibility within that as we develop our products, to ensure that we do it with an amount of integrity within there. We’re not always gonna get it right. We aren’t perfect. Sometimes we’re going to get things wrong, but it is, how do you just make that experience more fun? Sometimes that distraction, right? Like, as you were saying, going back and forth on the road, to just do those same intervals, because the data was steady, is not the thing that is going to get the best out of you. Sometimes you just need to switch your brain off and you think about something else. That’s why we all love group rides, right? I don’t worry about the power. I’m just thinking about the wheel in front of me. And I think, like augmented reality and virtual platforms do a great job of creating that experience where you’re riding along and next thing, you forget anything that you’re thinking about, and you’re just focused on that group and how to operate within the platform. I think where we need to be careful, because this has been seen within social media studies of the mental impact, the negative mental impact, especially on youth around that. And so I think it again, it’s important for us to be able to understand what our relationship is. And frankly, this applies to the whole conversation. Just because data is something that’s seen as purely objective for us or for a group of people, does not mean that’s how it works for everyone. There’s many cases where there’s negative impacts around even with HRV or step counters, they did a study where it was the number of steps that people were walking were either overestimated, accurate or underestimated, and the group where the steps were underestimated, so they were supposed to be walking 1000 steps, and they were told they walked 500 it actually led towards poor eating behavior, negative thinking, right? So it’s like you have to be careful around how these systems are put together, how you look at them. I think that you know social media is great when you can approach it. I like seeing it to where it’s like, this was a great ride. I love those pictures, right? Like seeing our friends and people that we know get to live their best lives and to peek a window into it. I think that is awesome. On the flip side, I think that when you compare. Yourself to that, and it starts to become something where I didn’t get to do that, or I’m not as able to do that thing. It can just have negative consequences. And it’s a understandable thing, right? Like, I felt that, right? Like I can’t say that I’ve never looked at someone’s great ride or someone’s power output and been like, Man, I wish I couldn’t do that. Right?
Trevor Connor 1:05:19
Like, of course I have, and that’s one of the concerns, is particularly when you have these segments where you are being ranked, you’re being compared to everybody else. And sometimes that can be hard on people.
Lee Gerakos 1:05:29
You can absolutely and I think you know that really comes through too when we compete, right? Like when you go and do a race, you get a placing. I think that if you focus too much on those placings personally, it can take away the joy from the experience. I think that’s one of the areas where gravel has really kind of taken off, is because VR just put all together, and it’s about like my own personal best, and I can have a great race regardless of where I end up, and I will connect with a group of people, and it’ll be fun to counteract that, right, like I’ve been spat out the back of a road race in the middle of nowhere, and it’s not as
Trevor Connor 1:06:03
fun, no. And I do think an important thing to point out here is we all know how negative the big social media platforms can get where you dispose one thing, and all of a sudden everybody’s telling you you’re Satan and the worst person on the face of the
Chris Case 1:06:19
planet has that happened to you? Trevor, what did you say?
Trevor Connor 1:06:22
Close like I described what I had for lunch that day, I will say, I’ve been on Strava for a long time. I have never gotten a negative comment. I have noticed that as a social media platform, it tends to be a lot more encouraging, which is probably a positive I would say, Yeah, I
Lee Gerakos 1:06:40
think there’s certainly positive experiences, right? Like, in being able to just be happy for each other, yeah, right? Like, I think that’s kind of what it comes down to. If that’s the experience that you get, and you have your own experience in your and you can enjoy those, and you can keep those separate and enjoy the experiences of others, it can be a wonderful technology and tool to bring people together. If that is difficult for a person, then it may not be that same experience, right? Like we don’t all have those same experiences. And I think that’s the challenging part. Is that, like, these are platforms used by many people, and we all have different ways in which we navigate them.
Trevor Connor 1:07:15
So the question I really wanted to ask you, because this must be something that you’re talking about, since you’re doing it. But up until now, there has been a separation of the gaming training platforms. You’re kind of Zwift, rgt, Ruby, Kino map, and the training platforms, like training peaks. Where do you see this going when the two are merged?
Lee Gerakos 1:07:37
Yeah. So I think in terms of, you know where we started, some of this too is kind of important. And again, we’ve grown with the technology, and we’ve grown with how cycling and coaching has evolved over time. As a new person coming into the ecosystem and working with a coach or getting a training plan, it’s very overwhelming, right? Like you need to get a device, you need to get a bike, you need to get a trainer, you need to connect all these things. You have different accounts. What is OAuth and like, what am I granting permissions to? And I saw it firsthand, and it was kind of like a moment for me where I was just like, wow, we take for granted just how much we have evolved with this industry. So really what we wanted to make sure is, like, how can we continue to just smooth that experience of people coming into the sport of cycling, right? Like, again, going back to all the devices and all the things we’ve been talking about, I need to measure how many things and what things just to ride a bike and ride faster. Like, is that the best thing for us? It’s super cool. And I’d love to geek out on the data. And I think that testing out new technologies and how your body interacts, right? Like, we’re all kind of like our own little science experience. Experiment in this process of, like, how can I continue to perform better? What works for me, what doesn’t? What are my breaking points? But that’s after you’ve been doing it for a while and you’re already hooked. I think there’s a part of you know, how do we make sure that people want to be doing this to start? And so how can we make that more approachable? So that was one piece of it. The other piece of it is that many coaches want to engage with their athletes. And again, even though we talk about like the human interaction, there’s also something wonderful about FaceTime and zoom and being able to connect online. And we do it all the time, right? Like day in and day out, there’s so much convenience to it, and we can still have fun and build relationships doing that. And again, for our support of the coach and for the credibility of the experience, what we wanted to make sure of is, how can we make sure that like that real world to virtual world starts to feel a bit more seamless? There’s kind of two aspects of it. There’s the training, the community, the events that, like any coach, any person, can actually go do and bring together. And then there is the credibility of the experience overall. And how can we make sure that for our customers, we’re helping them prepare for largely what they are doing, which is some sort of like either they’re serious about their virtual racing, which is kind of ties into some of the physics within the platform, or they’re preparing for some sort of real world experience. Right? And so how can we make sure that those two things as closely match as possible, such that you have real world feel, real world look, as well as now we’re introducing real world routes, right? So that is one of the things that we’re trying to make sure of. Is like, how can we just best prepare people for equally, we want to make sure that we have an open ecosystem, right? So we want to make sure that if that’s not the platform for you, great, but we want to make sure that that is an option for you, and we’re going to do our best to make sure that that is a compelling experience. We want to make sure that you can do your strength training, do any outdoor activity that you want to do, as well as support your indoor training. Right? Like those are those three big areas of training that we want to be able to make sure that our athletes can do.
Trevor Connor 1:10:41
As Lee points out, there are many ways that virtual platforms can be used. Here’s Matt sharp talking about the ways that he and his coach have taken advantage of them.
Speaker 2 1:10:51
My coach, Lance Watson, and through his company, life sport, I think they were one of the pioneers in online coaching. And so, yeah, he does an incredible job of, obviously, just checking in on me and using that kind of tech with life support. They do a lot of training sessions with spins and whatnot online, of course, which, yeah, helps build community. It’s fun to pop into those every once in a while and check in with other athletes and kind of, yeah, feel like the team atmosphere. And, you know, it’s funny, actually, I think about some of the stuff we did when he wasn’t with me during my Olympic build for a couple of training sessions on the track. Actually, we actually set up a tripod, you know, with my phone, I think we did like a zoom call, or set up a zoom call or something. So I was doing these intervals on the track, and basically he could watch me for the most part as I was going around. And I actually had, like, my AirPod in when I was running, so he was giving me feedback as I was doing these training sessions. So I think that was pretty to me. That was pretty innovative. I mean, he physically wasn’t in. Think I was in Arizona at the time, he was in Canada and so, but he was kind of on deck coaching me, which was pretty special, for sure. It was funny, though, because at the far end of the track, he’d kind of go radio silent because the Bluetooth didn’t reach that far. But, you know, you’d pop in, you know, on the other side of the track, or whatever, and it was kind of funny at the same time too. But yeah, definitely a great way to use technology and leverage it.
Lee Gerakos 1:12:10
You know, I know we’re talking about the virtual and the social part of it, but you know, you and I have chatted too to where I think there’s so much to training that is more than just always being on the bike. I’m always fascinated by Leadville every year, and one particular athlete, right, like we have the best mountain bike cyclists, especially within the US, compete every single year. And I think we’re all somewhat fascinated by John Gaston. Right, lives up in Aspen from the East Coast, I believe, like Connecticut, New England area comes out here as schema racer, professional schema racer, right? Looking to hopefully qualify for the Olympics. I know they have a competition here soon to determine that. But then does two mountain bike races a year. He does Aspen and then he comes to Leadville, and he comes in on the podium. This year, he was second, right? Like he was six hours and like 52 seconds. I think there’s a question of to me, would he be better if he was on the bike all the time, or is he as great as he is competing in that race because he’s not on the bike all the time? And I would kind of like pose that question back to you in terms of, like, how do you see coaching, athletic development and performance as it relates to other sports contributing to your overall ability on a bike?
Trevor Connor 1:13:20
Oh, that’s a good question. I love that question. I am a very big believer that if all you’re ever doing is being on the bike or whatever your sport is, if all you’re ever doing is that sport, you’re eventually going to get in trouble. I do think cross training, no matter what level athlete you are, is really important after that, I would say, I think there is big individual difference. I think there are some people that absolutely, I coach one athlete, he loves the bike, and I’ve tried every off season to get him out running, and he’s like, I hate running, I love the bike. So he just will do something different with the bike, but he wants to be on the bike where there are other athletes that I think if they did that much of a single sport. Example you just gave is probably a case of that. They probably wouldn’t be as good. It would just mentally cook them on the sport, and they’d be less interested where doing that big variety is actually what makes them stronger. I think that’s a very individual thing, right? So is that how you feel, or what’s
Lee Gerakos 1:14:18
your I think it’s all individualistic in that respect, it is what gets you excited and up in the morning and knowing again yourself. I think your athlete that loves to be on the bike all the time is a great example of that. I’ve done all these things. I just that’s not what I enjoy. Like being on the bike is what I live for. I think that’s a wonderful thing. And I think on the flip side, experimenting with that and knowing that is that thing and kind of like disqualifying those other options is also
Trevor Connor 1:14:44
great. Yeah, so I would say there, I think you brought up a good point. Self awareness is really key. When I talked to him about that, he had a really good answer that I went, I agree 100% with that. I’m not gonna argue with you on getting you off the bike, because he’s a CEO of a company, and he goes, Look, I have a stressful job. Job. I’m never going to be a professional cyclist. Cycling has always been my stress relief, my fun. So I get with professional athletes that need to have that off season. But when you tell me, in the off season, I have to stop doing the thing that is my break from what is stressful. I don’t like that. I want to be doing the thing I enjoy. And I went, that makes a ton of sense. That’s how you look at cycling. It is your break, it is your rest, it is what fills you back up mentally, right? So just keep you on the bike all year round. Yeah, I love that. Well, Lee, we’ve actually gone a little overtime here, but has been an absolutely fascinating conversation, and I know it’s a conversation that’s gonna happen again and again, because it’s all about where are we going, and I think the answers to all this in a year and five years and 10 years are going to be different, but great to have that check in and hear what you’re thinking right now. So before we dive into our take homes, we do have a question for the forum, and it’s really to all of our listeners just to continue this conversation. So just a very broad one. What trends would you like to see on the training platforms in the future? Would love to hear your responses. Love to hear what you’d like to see. And with that, Lee, I know this is first time on the show, but you’re a listener, so you know how we finish up. You’ve got one minute to tell us what you think is the most important thing for listeners to take from this episode, and do you want to go first, or do you want to finish us out?
Lee Gerakos 1:16:28
I’ll go first. Okay. Floor is yours. Thank you. What I would say in this whole conversation, what matters the most, regardless of what platform you want to use, whether you’re self coached, whether you want to use an AI platform, whether you want to work with a coach, I think is that the more people that we have in our sport, the more people that enjoy life, the more people that feel like they are bringing out their best selves. Everything is going to, you know, fall after that. And I think that you know, all the technology, all of the tools, all of the capabilities can help enhance that experience. And it’s not a one size fits all for every person. You know, we obviously have a particular perspective on what matters there. But just as you mentioned, your athlete that really wants to be on the bike, that’s what works for him, and that’s great. And other athletes, they need that downtime. Other athletes have the need for additional support, and other athletes just really enjoy doing this on their own and connecting with others. And I think that all of that is wonderful, and I think we should just be grateful that we all get to do this with our time. And I would just like to thank your listeners for listening to our conversation today, and thank you both for the time.
Trevor Connor 1:17:49
Absolutely, very kind to say the first person to ever thank the listener at the end, I appreciate that.
Chris Case 1:17:55
Chris, ooh, what I would take from this conversation is that we’re in a really interesting time to be an athlete, to be a coach and to be in the industry, supporting athletes and coaches, because things are changing rapidly, or seem to be there’s the potential for them to change very rapidly. But I think what we all have to keep in mind, both as users and as creators of products is that it’s not always the best to jump onto trends. It’s not always the most beneficial to just give in to the machines in a way like it takes a intelligent person to be discerning and make the right decisions about themselves as an athlete, or if you’re a coach, to make the decisions about the athletes you’re working with, because it comes down to individuals, and everybody’s different, and everybody has different interests, motivations and desires, and so it’s great that we’re talking about all of these things, because it will change. But it doesn’t have to be considered a bad thing. It doesn’t even have to be considered a good thing. It comes down to you, whether you want to use these new tools or not.
Trevor Connor 1:19:11
So I remember back 2004 2005 I had just bought my first power meter, and in order to look at any of my data. And I think this is also back wkO one. I had to take my heart rate data and my power meter data, put it into an excel sheet, and then line them up in the Excel sheet, export it to, I think, a GPX file at the time myself to pull it in and then have very rudimentary analysis of the data. So thinking about that compared to where we are now, it is light years difference. And I do think they’re accurate in saying the big trend that we just went through was that data integration and that ability to pull in a bunch of different data fairly easily and look at it all together right before this episode. Chris. You said, we’re kind of doing another episode on AI. We talked about this, and part of that’s because right now, when you’re talking about, where is all this going, AI is part of every little bit of that, I think that is the new trend, and I get that people are scared of it. I would say it has some amazing capabilities, but it also has some major flaws. So I don’t think it’s going to be taking over. I think it is a tool that is going to change and shift how we’re doing things, and we just don’t quite know yet how it’s going to have that impact, but it will have an impact, and the hope here is that it’s ultimately going to be a positive one. But I agree with both of you. At the end of the day, it’s people who have to interpret it’s the athlete and the coach who have the judgment that the AI doesn’t have. Yeah, well, Lee, thank you for being on the show. I’ve been excited to get you on, been working for what, a year now to get you on. So I really appreciate we finally got you here. Thank you both. Thank you. That was another episode of fast talk. The thoughts and opinions expressed in fast talk are those of the individual subscribe to fast talk wherever you prefer to find your favorite podcast. Don’t forget, we’re now on YouTube, as always, be sure to leave us a rating and a review. To learn more about this episode, from show notes to References, visit us at fast talklabs.com and join the conversation on our forum. Go to forums.fastlabs.com, for Lee gurakos, Dr Andy Pruitt, Ryan Ignatz, Jack Burke, Robbie, Ventura Matt sharp and Chris case. I’m Trevor Connor. Thanks for listening. You.
