Really appreciate you giving Vekta a try and for sharing this, these are totally fair points.
When someone joins Vekta, we automatically import up to five years of historical data from all integrations except Zwift. This is due to a current limitation in Zwift’s API, which doesn’t allow historical data to be passed over. I can assure you we’re working on a solution for this, as we know it’s frustrating for new users.
In the meantime, we have a support articleexplaining how to import your Zwift sessions manually, though we appreciate this can be time-consuming at this stage.
And you’re right about the app and desktop versions. The web platform has been our main focus since launch, so it currently offers more functionality. Over the next few months, the app will go through major updates to bring it closer in line with the desktop version and provide a more complete experience. Updates will start to roll out in the next few weeks.
Thanks a lot. What you are doing is quite courageous and seeing how old TrainingPeaks is and looks, Vekta’s more modern take on managing training is definitely a breath of fresh air.
Also, all my critiques, nitpicks and counterarguments come from a place of love. I’d like to see you succeed. Plus, I always wanted to live in the future. Product development takes a lot of time, patience and effort.
Another thing that is relevant: I have some experience with tools used in the manufacture of semiconductors, matrix headlights and other MEMS, etc. Think double-digit millions of €. A lot of the topics that you are encountering were also cropping up there, e. g. in the context of preventative maintenance. There are literally hundreds of sensor readouts, and most of them in most circumstances are a complete waste of time to look at — unless something bad/unexpected happens. Then you want those particular readouts surfaced to the user or tool admin. A simple example would be your car engine’s oil temperature: nobody cares if it is in the green range, only if it gets too high or if the engine hasn’t had time to warm up is it even relevant.
I have the impression we may have different ideas of what that means. From a coach’s perspective, what I’d like from a modern tool like Vekta is the following:
Take over all the busy work of training. E. g. I don’t want to count weeks to the A event when formulating training plans or patiently drop in “endurance ride”, “VO2max intervals” manually. That’s what computers are for!
Part of busy work is to decide on precise ramp rates for certain workout types, it’d be great if these were pre-populated with something sensible that I can then tweak as necessary.
I want to focus on higher-level tasks and not be distracted by a barrage metrics — unless I choose to.
Let me explain in a bit more detail what I mean. One analogy comes to mind is the difference in philosophy in cockpit design between Boeing and Airbus. Airbus has something called “dark cockpit”. When you look up in the cockpit a modern airliner, you have a “Christmas tree” of lights. Airbus’ philosophy is that all lights are off if the systems are functioning within normal parameters. This way if something goes wrong, you have a very easy time to find the affected system and deal with it. Boeing’s overhead controls in contrast are illuminated to indicate the status (e. g. “fuel pumps ON”).
When and how to get alerts is something supremely difficult with modern airliners as you have to carefully think about the relative importance of several concurrent alarms (if one of the engines is on fire, I don’t care that one of the lavatories is out of order).
Boeing’s older planes still have cables connected to the controls whereas Airbus’ side stick is completely digital. Boeing’s controls will directly convert pilot input to movements of the control surfaces whereas Airbus’ flight control system extracts the pilot’s intent and computes the appropriate control surface movements. Airbus’ flight control system (in the standard control law) only allows movements within the safe rage, i. e. it will not allow steep banking or very steep ascents or descents.
Vekta feels like following Boeing’s model and I would prefer Airbus’ model. And no matter the model, pilots (= coaches) need to learn the tool and understand how it works.
Here is what this means to me as far as a tool like Vekta is concerned:
Show green light if my athlete’s parameters are in the green, only show me a list of metrics if I choose to
What I’d like to see is something like that. If my athlete has slept well, fatigue, energy levels, HRV and any other metric is in the green, then all I want to see is a “green light” (or “no light”). Only if one or perhaps several metrics are off, I’d like to be alerted. The latter point is important, too: perhaps a single metric being off is not cause for concern, only if certain conditions are met do I get an alert.
When things are yellow or red, I’d like to have suggestions by Vekta on how to react. E. g. “make this an endurance workout” or “make this a rest day”. The choice is still mine, but clicking in “make this a rest day” would save me many clicks back and forth.
AI-Assisted Generation of Training Plans
The second thing I’d like to have is something akin to your LLM-powered training plan generators, where I, as the coach or (self-coached) athlete enter the relevant parameters and Vekta at least gives me a template. Perhaps I can choose templates (e. g. distinguishing only base and race (2 phases) or base, build and specialty (3 phases)). I should be able to further refine the template iteratively with words (e. g. “Shorten the specialty phase by a week.”), but the last thing I’d like to do is manually count weeks.
Suggesting Appropriate Performance Metrics for Given Goals
I’d like Vekta to be able to make suggestions for metrics given certain goals (e. g. I’d like my athlete to work on their punchiness). Finding good metrics is key and I think everybody would benefit from one another’s expertise, i. e. if “secret sauces” were shared.
Now people will point me intervals.icu, but my reaction is decision paralysis: I can compute anything I want from the data, provided it is tracked somehow, but I am still no closer to knowing how to quantitatively measure “durability”.
IMHO if you are able to develop not only a good library of metrics, but suggest them to coaches after specifying their goals, you’d have a winner on your hands. It’d be great if you made “the best kept secrets” of coaches available to anyone — and thus lay the foundation for the next round of improvements.
AI-Assisted Generation of Training Plans
Now I know that you stated that this isn’t on your agenda at this point, I think this would be a key enabler for coaches. As a coach I’d like to decide how hard to push an athlete (or to give them time to rest when the situation calls for it), and then be presented with options that implement that. E. g. I could decide that my athlete needs two VO2max workouts per week and decide what the expected progression is (think of an “aggressiveness” slider). But then I can deviate from that by e. g. making it easier or harder (perhaps have variants of a workout ranging from -2 to +2).
I think that is understandable, but let me argue for allowing the use of several platforms. I have been using TrainerRoad for over 7 years now, I think. When I had time to race, I would subscribe to TrainingPeaks to be able to analyze races and outdoor rides.
On the other hand, I really didn’t want to do all my training in Training Peaks as I felt their model is outdated.
Vekta sits in between, which is a really dangerous place to be. I’m sure it works great for professionals and other coached athletes, but for Vekta to be a viable business, it IMHO has to expand to self-coached athletes. For self-coached athletes, the app needs to address both roles (that as coach and that as athlete). That’ll make app design challenging, but I think the payoff will be worth it.
E. g. if Vekta had great race analysis tools, readiness scores and performance analysis/target metrics tools, I’d maybe want to use it in addition to, say, TrainerRoad.
Anyway, I hope I am not coming across too negatively or am completely missing something obvious, because chances are, you probably have thought of it, but need a few years to implement it properly.
I didn’t really understand this statement. If I understand correctly, Vekta doesn’t automatically create workouts or plans and doesn’t have a workout player. It feels much more like modern competition for tools like TrainingPeaks, which I don’t think anyone would call a complete ecosystem. I think it’s an interesting concept, but the marketing side is confusing. The fact this thread started with thinking it was a new competitor for TrainerRoad says a lot, as does the difficulty in being able to simply explain what the tool is and does and who would benefit from using it.
I don’t understand what is being said here either. It reads to me like “Vekta doesn’t create training plans or sessions automatically, you have to build them yourself with our session builder and calendar’, but on your front page, it says
Wow, that’s a much better way to succinctly explain my confusion about what the product is actually meant to do.
TP is designed to work also with other platforms, so it is lacking some features like a workout player. On the other hand, if Vekta is designed to be used stand-alone, then it misses basic features like the ones you listed. A workout player for indoor workouts would IMHO be part of a minimum viable product feature list, for instance.
I mean it is clear that Vekta is a newly developed product and we cannot expect that it is released with all the features. However, the vision @Dominic_Valerio shared leaves me a bit concerned, at least if Vekta is meant as a product for a broader audience, meaning athletes who are self-coached.
Expectation management and proper marketing is crucial. If I have to create a plan by hand for a season, that’s not done in seconds (unless I copy-and-paste — if that is an option).
I was thinking on this more and it occurred to me that I was a bit overly critical of TP too. In TP you can buy Training Plans and use TPV to do workouts, so while it’s not “smart”, or pretty, it is, in fact, a complete ecosystem. Plan, train, and analyze.
Let me think for a nanosecond…