At this point I have to assume they’ve given up. The radio silence every time it’s brought up has led me to that conclusion. Does someone from TR care to confirm or deny that?
Yeah he’s been a lot less publicly active with TR in general for quite a while, besides a brief stint of PR for the Zwift integration. Looking at his Strava he’s not been riding much either. Seems like he’s lost his passion for cycling.
TR is clearly no longer Nate’s hyperfocus.
Which is a shame, as his ADHDesque enthusiasm was infectious. I loved some of the early podcasts with him, Chad and Jonathon bouncing off each other.
I’d be careful to not jump to conclusions. I’d focus more on what TR-the-company does as opposed to how much we see of @Nate_Pearson in public. Feature updates have steadily come out over the last few years, including periods when @Nate_Pearson has been quiet. Two very recent features that come to mind are Red Light/Green Light and volume recommendation in Plan Builder. @Jonathan has been steadily hinting at additional features.
There might be plenty of reasons why Nate has been less public, and the reasons need not be (all) bad. I reckon the overly optimistic estimates when it comes to WLv2 might be one reason he has shut up about upcoming features until they are ready for at least alpha/beta testing.
That might be true but as pointed out it’s been two years since WLv2 has even been acknowledged as an active project. It would be nice to know if it’s still been worked on or not. The fact that this question keeps getting asked and TR doesn’t even give a vague reply concerns me.
Edit: I did find a more recent TR comment addressing assigning PLs for non-TR workouts/activities. Here, and then reaffirmed a few posts later in June.
Instead of event when building a plan I’d like to be able to say for the next X weeks build a plan that focuses on Y durations.
So I know at my current FTP my z2 stuff is good enough for me but 10 minute efforts are terrible so I’d like a plan to focus on that. Could break the durations in Short <5 mins, medium 5 to 20mins and long >20mins.
Could have a wider range of durations or just medium and short with longer ones included in normal plans.
Don’t get me wrong, I want to know what’s coming up and want new features yesterday. I also liked @Nate_Pearson ‘s way of being somewhat open about TR’s development. It’s not that I prefer the “new way” (if there is such a thing), I’m just saying that it doesn’t necessarily mean TR has stopped evolving.
However, the focus on WLv2 and features like it may lead TR to neglect resolving “boring” issues like
I could add more to this list, but I think you get my point.
Nate’s newfound interest in strength training gives me hope that strength training and other sports are added as first-class citizens. Think of how TrainingPeaks evolved from a tool for cyclists to a tool for coaches and athletes of all stripes, it’d be great if TR could add a few of the most popular (endurance) sports and strength training.
The thing is that feature development feels soooo slow with TR. There is no open roadmap or “uservoice” like suggestion ranking and all you get is endless “under the hood improvements” version updates.
Additionally features are becoming a bit inconsistent:
Over the last years it’s become less and less fun keeping my subscription (been here like 7 years). Always stayed for the nostalgia and hoping that next year would bring some real WLv2 breakthrough with a consistent UI but that never happened for me.
I’m not on a plan any more…more like self coached and mainly use TR as a workout player. Seeing the fast development of intervals.icu (what a contrast to TR) I could see myself just using that in combination with a minimal workout player (my Garmin or TrainerDay) in the future…
I’m not sure if it’s pure strategic decision or more like personal / health related but I miss the old Nate on the forum (the time before the breakup with Chad).
There very much should be (and almost certainly are, due to the different required skill sets) separate teams working on front end software (apps, website) versus backend software (all the data processing, training plan generation, AT, AI, etc). So work on WLv2 and improving the apps aren’t competing for the same resources.
I agree with @Rizzi. The development has felt super slow for a while now. Not just WLv2, just general progress, especially on fixing obvious issues like various features that are app only or desktop website only. As a user at least being able to do everything on mobile (even if the app kicks me to the mobile website) feels like minimum required functionality
I get that a lot of work (front and backend) went into re-doing the calendar. And we did get some nice new little features there. But again that’s still not fully functional on mobile. And IMO calendar improvements alone aren’t enough progress for the subscription prices being charged.
Part of that may be perception. There’s lots of little tweaks but as Rizzi said they aren’t really graphically seen by the customer, not all that clearly communicated.
Hopefully, TR has been doing some major development and it’s all tied together and we’ll see a huge enhancement drop. Sometimes there’s a lot of tech debt or a new feature requires re-doing a lot of code due to a new approach.
And I can understand if that isn’t talked about until the launch of it. But that’s the risk of subscriptions. Users are paying every month or year. I expect to see consistent signs of improvement if the price of the subscription is higher than my perceived cost for TR to run the existing service profitably.
They are competing for the same resources: money and the time devoted to them by management. With the former I mean that it is very easy for shiny/sexy-new-feature to take attention and money away from the “boring stuff”. E. g. management hires someone to work on the sexy new features rather than look for an app dev.
If you add “development of the apps” to your paragraph, I completely agree. With the ML/AI stuff as well as training plans, I do have the feeling that they have been making the most consistent and best work on it. Here is my experience regarding the app and the web app:
Work on training plans most certainly includes lots of changes under the hood that were invisible to regular athletes:
New features:
We can argue what constitutes slow/fast, but let me use a more neutral phrasing: I think app and web app development is TR’s weakest side, TR has become quite stale and there are IMHO quite a few gaps in functionality. Basically, you at least need to use two clients to have access to full functionality, and IMHO that’s a mistake.
What adds to the perception of staleness of the app and the website is that the app’s design hasn’t appreciably changed since I started using it. The Workout screen is nigh identical, unless you count extra options in the advanced menus. Fairly or not, this leaves an impression with the user. To be clear, I don’t clamor for a redesign of e. g. the Workout screen, just more functionality, though.
Iterations on training plans and training plan-related features are less visible, because athletes “don’t see versions” and we don’t know what the changes precisely are as our training plans are all different anyway. Still, in these areas, I do see evidence of improvements. And I see new ML-/AI-based features added, although probably we should stop caring that they are enabled by Machine Learning or “AI”. They simply are features.
I’m also eagerly waiting for WLv2 because my workout levels just deteriorate with the swapping to group/solo rides and it feels like I’m starting again from the beginning come autumn. That being said, the AIFTP taking the outdoor rides into account helps a fair bit.
This is true, but man the release of Zwift integration taught me a lesson about how great the TR workout player is. I instantly feared that Zwift was going to acquire TR and kill the app.
I agree with most of your post, but I don’t think many care about these two features. I’m an all Apple household and these are low priority for me. You can use HealthFIT to sync your data from many sources to/from Apple Health. And I don’t believe Apple even provides a good way to transmit heart rate data at a 1s data rate to other apps. What percentage of TR users really care about these? Maybe 10%. What percentage want non-TR rides to contribute to PLs? I’d wager something like 75%.
They did do one big update that I forgot about: Zwift integration. It’s not something I use, but loss of users do.
There are so many reasons why you’d want that, and basically everyone, and that’s the inclusion of sleep data in the Adaptive Training algorithms. Sleep is one of the biggest determining factors in how well your workout will go.
Also, I don’t think HealthFIT can sync e. g. body weight and total calories. Supporting the platforms’ standards just seems logical to me. Apple Health should not be treated differently than Garmin’s or Wahoo’s clouds.
I forgot about that, thanks for reminding me. I have never used Zwift and have no desire to, but this was an important feature update.
It’s great, but I can think of a few ways to improve it, e. g. allow for customization and more data fields on, say, iPads.
Not to mention that sleep data (at least on Garmin, can’t speak to other devices) is questionable at best, GIGO.
In the smartwatch market, Garmin is a tiny blip. And it doesn’t help if you are outside of the Garmin universe (partially or as a whole). So the size of the data pool would be much, much smaller.
That’s because they likely can’t. They would need scale to make sleep data useful. IMHO it is a big weakness in their strategy, they can’t even try a lot of things with statistical significance without it.
Sleep tracking seems to be a solved problem with the Apple Watch, at least when it comes to sleep duration. It gets trickier when you want to know about sleep phases.
The counter argument to this is that you don’t need to use the data immediately, and the sooner you start aggregating it the easier it becomes to use it at a later date
that said, I don’t think Apple health integration should be prioritized
There are plenty of competitors to Garmin sports watches (Polar, Coros, etc.), none of which to my knowledge can put their data into Garmin’s cloud. I am one of the 5 humans on planet Earth with a Wahoo Rival, so I reckon TR could access my sleep data (if it hadn’t stopped working … again).
Even if non-Apple/non-Android wear smart watches are overrepresented amongst TR athletes, I still reckon that Apple Watch still retains the highest market share.
I bet it’s not even that close.
Also, maybe I’m misunderstanding the situation, but I think other devices who implement HealthKit features can also maintain sleep data in the Apple ecosystem.
The thing is, just while TR was speaking about WLv2 there have been complete new AI apps left and right entering the market (enduro, join, aiendurance, humango…). Of course we can argue about quality and feature set but that contrast alone just gives a feel of slowness for TR, especially considering TR premium price tag.
Also when you see the what just one dev can do with intervals.icu (ok meanwhile David has a supporting UI dev), it’s crazy.
If it wasn’t for DC Rainmaker having TR in second position after Zwift…I’d assume TR is in trouble.
Forum and podcast have become less and less fun over the years. For me the 75% reason is Nates vibe is missing and the other 25% is other key players gone (Chad, Amber, Pete, Ivy). The funny thing is IIRC Jonathan said 1 or 2 years ago that podcast numbers wouldn’t reflect that (I can only assume it’s because the top podcasts as a whole are growing and TR profits from a well deserved pole position in that spot…but what do I know…maybe I’m just getting old 
)
I think quality is really important: it is easy to announce something that doesn’t work right. I remember FasCat Coaching announcing xPower/xFTP/not-sure-what-it-is-called and it was validated on 5 athletes and 100 workouts. When reading about e. g. Garmin’s training readiness scores and features like it, it seems like a mixed bag, something of rather limited utility. Doing something right takes a lot of work and makes it appear as if you are doing less, especially if what you get is a very simple signal (think RL/GL) and further advances are invisible to users (think AT or RL/GL again).
Apples and bowling balls: TR is a company with at least around 100 employees (that’s the last number I remember), which means it is an entirely different thing. Designing AI-/ML-based features is very, very hard as those devs are hard to come by and those who have some experience in exercise physiology are rarer still.
From my own experience, it is all the boring stuff (automatically cleaning and treating datasets, etc.) that takes a very long time to get right. It isn’t something that can be done by a single person for sure.
Yup, I agree with you on the podcast, it has become quite sterile and is missing the warmth and banter. I would also appreciate some semi-regular by @Nate_Pearson on what’s to come and what their thinking was that went into designing features. The current format is informative, yes, but not as fun.