First of all, I’d like to thank the entire Athletica team for building such an ambitious platform. I originally tried Athletica more than 1.5 years ago, subscribed for six months in 2025, and recently came back for another six-month subscription because I genuinely believe in the philosophy behind the platform.
The science it is based on is solid, and I don’t know of any other platform that offers what Athletica does. The physiological modeling, automatic threshold detection, AI coach, Workout Wizard, and individualized training are exactly what I was looking for.
However, after using the platform for well over a year in total, I still feel that Athletica is held back by many small bugs, inconsistencies, and unfinished features that constantly distract from what should be the main focus: training.
1. Long-standing issues remain unresolved
My biggest frustration isn’t that Athletica has bugs—every software platform does. It’s that some of the same core issues I encountered during my trial more than 1.5 years ago still exist today, despite being discussed multiple times in the forum and with support.
For example:
-
When using Stryd, running pace is still imported incorrectly, even though the pace data is correct in Stryd, Strava, and every other platform I use. This makes pace-based AI feedback on time in spent in zones unreliable and frustrating.
-
Heart rate zones can only be anchored to Critical HR. There is currently no way to define zones using separate physiological markers, even though many athletes have access to this information through laboratory testing. As a result, you have to choose between having realistic low-intensity zones or realistic high-intensity zones—you can’t have both.
-
Session feedback often evaluates workouts using metrics that were never intended to be the primary target. For example, VO₂max sessions prescribed and executed perfectly by power are criticized because of heart rate or pace compliance, while easy aerobic sessions can receive comments based on pace or power instead of heart rate, which is the target metric the workout should be based on.
-
Some workout targets themselves don’t seem physiologically realistic. For example, a 30/30 VO₂max session expected almost all of the heart rate time to be spent in Zones 1 and 2. Even though power should be the primary pacing metric for these workouts, the heart rate targets still don’t reflect what actually happens during this type of session.
-
Switching the preferred running metric to power, only for it to automatically switch back to pace a few minutes later. You can not get Athletica to work with running power as the preferred metric.
-
The AI coach sometimes hallucinates or changes its answers when you continue asking questions. This means you can’t always trust the responses it gives, and they are not always consistent with the Athletica handbook.
None of these issues individually would stop me from using Athletica. The problem is that together they create constant friction. Instead of trusting the analysis, I often find myself wondering whether the graphs, zones, workout targets, or AI feedback are actually correct.
Considering that several of these topics have been discussed for well over a year, I would personally prioritize fixing these long-standing issues over introducing new features. A more polished and consistent experience would add far more value than expanding the feature set.
2. AI coach feedback should be more context-aware
The AI coach is one of Athletica’s unique selling points, but at the moment its feedback often feels disconnected from the purpose of the workout.
For example:
- VO₂max sessions prescribed and executed by power receive negative feedback because of heart rate or pace.
- Easy aerobic sessions can receive comments about pace or power even though the workout was intentionally prescribed by heart rate.
In my opinion, the AI coach should primarily evaluate the metric that the workout was prescribed with. Secondary metrics can certainly be mentioned, but they shouldn’t dominate the feedback or suggest poor execution when the primary target was achieved.
3. More flexibility
Personally, I don’t always want to follow a strict day-by-day training plan.
My main reason for using Athletica is actually the analytics:
- Power-duration modeling for running and cycling
- Automatic detection of physiological markers
- Individualized training zones
- Workout Reserve
- Long-term performance tracking
Especially during the summer, I do many long social rides where the exact structure isn’t important. I’d love to have a “free training” mode where I simply perform my activities, let Athletica analyze them, update my physiological markers, and provide insights without trying to optimize every day according to a predefined training plan.
Especially when setting new power PRs, I love Athletica’s analysis—seeing Workout Reserve drop below zero, knowing I’ve pushed my limits further than ever before, and then seeing my power-duration curve and training zones update as a result. Knowing that the intensity target for my next 30/30 workout is tailored perfectly to my current physiology is where Athletica really shines for me.
4. Better support for hybrid athletes
Many endurance athletes don’t fit neatly into a single sport. For example, my primary focus is running, but I also race bikes, commute by bike almost every day, include strength training and I’ve done one Hyrox competition.
Currently, there isn’t a clean solution for this. The recommended workaround is using a duathlon plan, but that requires creating fictional duathlon races, regularly deleting brick runs, and other annoying steps that just waste time.
I’d love to see a more flexible planning system where athletes can combine running, cycling, strength training, commuting, and occasional races without relying on workarounds.
Commuting, in particular, is very cumbersome in Athletica, as you’ll always receive negative feedback about unplanned sessions unless you manually create a workout for every commute in advance.
5. Apple ecosystem support
Athletica works best with Garmin. Using an Apple Watch requires several workarounds and even additional paid apps such as Watchletic to obtain the same functionality.
A smoother Apple ecosystem integration—especially for HRV, resting heart rate, sleep, recovery data, and workout syncing—would make the platform much more attractive for Apple users.
6. Overall user experience
Most of the individual issues I’ve mentioned aren’t deal-breakers on their own. The problem is that they accumulate. Incorrect pace analysis, inflexible training zones, workouts occasionally failing to sync, AI feedback based on the wrong metrics, and other small bugs create constant friction throughout the user experience.Instead of thinking about training, I often find myself thinking about the software.
Final thoughts
Despite all of the above, I renewed my subscription because I genuinely believe Athletica has enormous potential. I haven’t found another platform that combines physiological modeling, threshold detection, training prescription, and AI coaching in such an interesting way.
That’s exactly why I’m taking the time to write this feedback.
I would love to see a stronger focus on polishing the fundamentals before adding more features. Fixing long-standing bugs, improving consistency across the platform, making the AI coach more context-aware, and providing more flexibility for social rides and commutes would dramatically improve the day-to-day experience.
.