I’ve used all three, along with Mottiv, and TrainerRoad.
@Dazza has done a good job of summarizing most of the pros and cons, TriDot also syncs with Form goggles now as well btw.
Look, in the end, you are putting your trust in an algorithm/AI, which, especially in the case of AI, quite often even the programmers don’t know exactly what is happening. LLM models are literally a black box that no one really knows what is going on inside of.
But Athletica is the only one that explains the science behind it’s choices.
TriDot “adaptations” I am convinced, after using for over 6 months before coming to Athletica, are either so minimal it’s hard to notice, or just plain lies. You are never, ever, told what adapted.
Don’t even get me started on their touted Physiogenomix (you upload your genetic test from something like Ancestry.com and it’s supposed to adapt accordingly). I did that as well, it made … zero difference to the training plans and just seemed like a big gimmick.
I do like that TriDot (supposedly), use weather/heat data from sessions, and I think sleep data as well, for their planning (not that you would ever know, cause they never tell you what changes). I wish Athletica would do that as well.
TriDot is slick looking, and did a good job from getting me from 0% to 50% of where I wanted to be, but after that, it seemed to be content on keeping me at mediocre for way too long, even when I moved to their high volume plan. And again, zero explanation of the science.
Humango (which I used before TriDot), had potential. Hugo, it’s AI robot was laughably bad and pretty much useless. It did have some nice features, but I just continually hit bugs with it over and over again. Now granted, Athletica has its fair share of bugs still as well, but the response time is far better and way more transparent.
Their interface looks “ok”, but is clunky and overcrowded I found.
Motivv isn’t a bad choice for beginners, not a lot of smarts, but solid enough. It is essentially developed by one main developer, but I did like it’s nutrition advice for each session. Would love to see that brought into Athletica. Never felt like it was advanced enough though once I moved to higher training loads.
TrainerRoad is still way too cycling focussed to be of real benefit to Triathletes. There are some similarities between it and Athletica with its adaptations and it’s red light/green light system (essentially like Athletica’s warning triangles on sessions), and it’s interface is no doubt slicker at the moment, but it doesn’t adapt swims, nor do red light/green light on those, and doesn’t push out to Garmin for some things.
If you listen to their podcast with the founders, you’ll realize they have no intention of focussing on Triathlon anytime soon either.
In the end, my view on all the platforms, is to use the AI smarts for adaptation suggestions, but use my own brain in the end for final decisions. But to do that, you need to know WHY the adaption was made based on what foundation, and that is where Athletica wins by far.