I’ve been vocal about my concerns with the latest version of TrainerRoad, but those concerns come from wanting the platform to succeed. TrainerRoad has been an important part of my training, and I still believe it can be the excellent coaching platform I once considered it to be. Currently, however, it functions more as a workout library, as I’ve modified roughly 9 out of 10 prescribed workouts over the past several months because they were significantly easier than expected, regardless of workout ratings, increased targets, or reduced recovery periods.
Rather than viewing forum criticism as negativity, I hope the team sees it as valuable feedback from engaged users who want the product to improve.
I’d encourage them to review my training data as a case study, particularly my outdoor rides and how the system interprets them, as it may provide insight into why its recommendations have missed the mark for the past six months. I’m happy to provide additional context or answer questions; my goal isn’t to complain, but to help TrainerRoad become as effective as possible for the athletes who rely on it.
The bad news is that, from my perspective, TrainerRoad has been almost completely ineffective for the past six months. Frankly, if I were on a month-to-month subscription, I would have paused it by now.
Throughout this period, I’ve effectively had to self-coach every week and manually adjust every intensity workout based on what I’ve learned from previous training blocks. I haven’t felt confident following the prescribed progression without significant intervention on my part.
The good news is that my upcoming intensity week appears to be the first one this year that follows a logical progression from the previous intensity block. For the first time in 2026, I haven’t adjusted the next two workouts, which is encouraging. Hopefully they won’t be significantly reduced again after the recovery week.
At this point, I don’t feel like I’ve received much value beyond access to a workout library and calendar, resources I effectively already had available offline. That said, I do appreciate the recent changes I’m seeing, and I’m hopeful this is a sign that the platform is moving in the right direction.
I think I speak for some of your customer base when I say; it feels like we, as customers, have an unusual relationship/loyalty to TrainerRoad as a brand. Most likely due to our proximity to the founders and employees. At times it feels like a family, and ultimately, we’re rooting for you.
Train your gut. Really know how many carbs/hr you can handle. If you’ve been training with 60g/hr and come race day start throwing 90+g/hr at it hour after hour your stomach revolt
Sort out your hydration needs and plans. I’ve started drinking every 10 minutes. If I don’t do that I’ll drink fare too little even if it’s hot.
If you fall behind on your nutrition during a race trying to play catchup and throwing a bunch of carbs at once may once again cause your stomach to revolt
Figure out your cold and hot weather nutrition plan. If it’s hot and you’re chugging bottles that have too many carbs maybe you separate your hydration from your carbs. Or just have bottles for water and then concentrated carb bottles you sip from.
I may try separating carbs from hydration. I recall Alex Wild mentioning he does that and just takes a Carbs Fuel Gel every 30 minutes.
And on hot days plain water can be refreshing.
Today I had a mild weather day and did 90g carbs per bottle in 3 bottles. Taking sips every 10 minutes. It worked out perfectly.
For something as long as unbound I might just do water with some sodium and then use gels and some “real” food.
None of these things are revelations but it’s more just the point of you have to take nutrition seriously and get it dialed in for what works for you. Some people have iron guts and it doesn’t matter. For me, my gut will shut down if I over do it. And in a long race like unbound, 100 miles in, and it’s 90+ degrees you do not want your gut malfunctioning.
I could have written that exact post. Exact same experience over the same timeframe. Also just using TR as a calendar. My next race js about 3 weeks away and I just copied what I did leading up to it when I last had a good result. Ignoring predicted and detected FTP and going back to FTP testing.
Hoping TR gets back on track with respect to growing fitness again.
Couldn’t finish this one as the chest thumping got too deep. Too many conflicts of methods and too much bragging to stay with it…Please go back to informative, research based topics.
ive noticed I have very few longer rides on my schedule leading into leadville. Im not really unhappy with this. This is my 3rd attempt and Ive not made my cut offs on the past ( 10 min, 4 min misses) I spent the first summer doing weekly centuries, the second year again several long rides. This year my longest was unbound and now i am back mostly 2-3 yr rides. I feel by far more prepared doing the 90 minute power works and actually hitting my numbers then when i just did lots of long grueling stuff. I am cautiously hopeful.
I think it is also that I feel their incentives are aligned with those of athletes: they make money by a subscription and they don’t sell our data to anyone. They want to keep us subscribed by making us faster by following the data and the science.
Like you wrote that doesn’t mean they don’t make mistakes or ship products with (significant) bugs. But from my interactions with their staff I know they care.
Moreover, I don’t see any competitors that come close. A good human coach might be better, but then we are talking at least 10x in cost. And when you pay that much, your expectations should change.
I see TR as the most “clinical” of apps which is one of its big appeals for many.
Some other apps take a slightly more “holistic” approach which appeals to some.
I had a coach for many years. It was definitely nice just following whatever showed up in TrainingPeaks and feeling confident you’d be ready race day. I think the reassurance from a coach regarding proper rest was always a big plus.
TR is very data-driven and I think this is one of the places where it shows. If you think about it, it is very hard to design a product for all personality types at the same time. A coach can adjust his feedback to the personality of the athlete, but the version of TR that you use is the exact same one that I use. It has to appeal to data-driven athletes as well as to those that don’t like to delve into the numbers, for instance. To some showing FTP predictions is helpful, to others it is a source of stress.
In episode 590, Jonathan mentioned that he might try some form of hypoxic training using a mask/device, though he hadn’t bought it yet. Curious as to which product?
If he does end up experimenting with it, I’d really love to hear an in-depth discussion on the podcast about the process: what device he chose, how he used it, how he incorporated it into training, and whether he felt it provided any measurable or noticeable benefit.
I tried something similar back in the early 2000s, based on the kind of device he seemed to be describing, and I honestly don’t know that it made any meaningful difference to my performance. So I’m curious what research or evidence he has read that makes him interested in trying it now.
It would be especially interesting to hear whether he sees it as potentially useful for endurance adaptations/altitude simulation
There’s the “set it and forget it” crowd who set their plan, do the workouts, and just don’t worry too much about the details unless things are way off the rails.
I think there are some deep diving data nerds who question every workout and every plan and turn every knob in TR to see what happens. Definitely not the “just follow the plan” type. These probably are the people who help root out a lot of the bugs or issues I think.
Then there’s probably a lot of people in the in between zone. Mostly just follow the plan but maybe dig a bit into the data. I think I’m in this zone. I mostly want TR to do its thing. I’ll adjust workouts as needed but I’m not motivated enough to go deep diving into a lot of optimizations around my training.
I really like discussions like this, because you are not just talking about a technical feature, but how to make them usable and useful to athletes.
If I zoom out, a lot of TR’s problems is that their new system has quite a bit of added complexity, which some users really want and others don’t even know exist. E. g. if I were in TR’s product design meetings, I’d advocate against surfacing FTP predictions the way they are used now. Because some people will fixate on that, see some fluctuation (“I rated my workout Hard and now my FTP prediction decreased by 3 W, what is going on?!”) in the FTP prediction, which may be due to stastistical noise essentially, can really lead people to overthink thinks. Or they may see this as a “promise”.
Even if we factor out accuracy, seeing the FTP prediction could be very motivating to some or very demotivating to others (“What, ‘only’ 299 W after all this work?”).
My perspective - take this with a massive grain of salt as I dropped my legacy priced TR subscription over the changes to AI FTP DETECTION - is that TR did its self a massive disservice with showing FTP Prediction by default and not creating a mechanism to show / hide this completely - not the current you can hide one FTP Prediction, but it comes back in the future. Having FTP Prediction become such a prominent part of the UI/UX without a means of not showing it implies to the end-user that the only thing that matters in training is improving their AI FTP number. And that if the number isn’t going up, then their training is a failure.
This single minded focus in the UI / UX on increasing AI FTP goes against a lot of what TR has been emphasizing in podcasts / other material that training isn’t just about increasing FTP. And that increasing FTP isn’t the only way to get better / stronger.
An athlete’s current and predicted AI FTP are not the only two metrics that are shown, though, even though I agree they are by far the most prominent: they also show a power-duration curve to show e. g. how the currently selected workout stacks up with other recent efforts.
Secondly, apart from FTP, there is no good, universal and easy-to-understand metric to measure performance I am aware of. In specific situations, you can come up with some, but these will be very personalized.
For the vast majority of TR’s athletes, FTP correlates very well with essentially all other metrics you can think of, simply, because there are a lot of gains to be had. They lie on the steep part of the S curve. For someone like me who is close to maxed out and gains are small, it is less good, though.
It’d be great if they offered more performance metrics, but honestly, this is likely quite hard. TR’s app has gotten more complicated already and I imagine that people who are new to structured training might easily get overwhelmed already.
