I’m currently using TrainerRoad alongside a coach, and I’d like to better understand how AI FTP Detection and Adaptive Training interpret workouts when I manually reduce the workout intensity.
For context:
My coach currently uses 318 W as my working FTP.
TrainerRoad currently estimates me around 331 W, with a projected/target FTP around 337–342 W.
I do a lot of structured training outside TR as well: TTT, Zwift races/ZRL-style events, endurance rides, and strength work.
My concern is that if I let TR increase my FTP too aggressively, the prescribed TR workouts may create more fatigue than intended by my coach’s plan.
So I’m considering keeping AI FTP Detection active, but when TR sets a higher FTP, using workout intensity at around 94–96% to keep the actual targets closer to my coach’s intended workload.
My questions are:
If I reduce a workout to, for example, 95% intensity, does TrainerRoad clearly account for that when evaluating the workout?
When I complete the post-workout survey, is my response interpreted relative to the workout I actually performed at 95%, or relative to the original 100% prescription?
Does frequently completing workouts at 94–96% intensity affect AI FTP Detection or Adaptive Training in a specific way?
Would it be better to manually set my FTP lower, or keep AI FTP Detection active and adjust workout intensity on a case-by-case basis?
Is there any downside to using workout intensity adjustments as a way to align TR workouts with an external coach’s FTP/load prescription?
I’m not trying to “game” the system — I just want to keep TR useful while avoiding unintended extra fatigue on top of my coach’s plan.
Thanks for any insight, especially if anyone from TrainerRoad can clarify how intensity adjustments and surveys are handled internally.
You’ve got lots of good questions here! Let me break them down in the order that you provided above..
Yes, we’ll analyze the data you record with your power source, so we’ll know where your power is in comparison to the targets.
Answer the surveys based on the work you did, not what was prescribed.
Not necessarily. If you’re doing workouts with a lowered intensity and they still feel like they’re appropriate for your abilities, we’ll take that into consideration when prescribing your upcoming workouts.
It depends on your goals. We usually recommend keeping your FTP wherever TR has set it, but some athletes have manually adjusted their FTP a bit lower and have enjoyed getting higher-level workouts with longer sustained intervals. Regardless of where you FTP is set, you’ll still get workouts from TR that are appropriate for your abilities. With a lower FTP, we’ll just give you higher-level workouts.
This is where things get tricky. What is your coach prescribing? How does that align with what TR is scheduling for you? I’m not sure I understand how you’re utilizing both, so it might help to hear a little bit about that. Either you’re following TR’s workouts, or you’re not, and that does make a difference.
Eddie has it on all the answers but I would also add, it depends what your human coach is doing and why.
If you/the coach have completely different objective, then don’t use plans and let the coach prescribe workouts.
If you bike training is aligned to a TR plan and the coach just wants to tweak it based on their interactions with you and knowledge of your non TR training, then yes your intensity tweaks would seem the right way to go.
The clarification that I should answer the surveys based on the work actually done, not what was prescribed, is exactly what I needed.
To give a bit more context on how I’m using both TR and my coach:
I’m not trying to run two independent training plans at the same time. My coach is the one managing my overall training load and prescribing my key workouts in Nolio. His main objective for me is aligned with what I’m targeting in TR: L’Étape du Tour 2026, while also keeping some focus on ZRL / Zwift racing and TTTs.
The reason I still use TrainerRoad is mainly because:
I really dislike traditional FTP tests, so AI FTP Detection is very valuable to me as a trend/validation tool.
On easier endurance days, or for example when I need around 60 minutes after a TTT, I often use TR workouts because the formats are convenient and well structured.
My coach often prescribes workouts that are inspired by TR-style sessions, but they are frequently adjusted and sometimes harder depending on the wider context of my week.
My concern is not that TR is wrong. It is more that if AI FTP Detection increases my FTP faster than the FTP my coach is currently using for prescription, then executing TR workouts at 100% could unintentionally add more fatigue than my coach intended.
So my practical idea was:
keep AI FTP Detection active as a useful signal;
keep my coach responsible for the overall load;
when using TR workouts, adjust intensity down slightly if needed so the actual targets remain aligned with my coach’s current prescription and the rest of the week.
Based on your answers, it sounds like using intensity adjustments can be a reasonable approach in this hybrid setup, as long as I answer the surveys based on the workout I actually completed and not the original prescription.
In other words, I’m not trying to “game” Adaptive Training — I’m trying to keep AI FTP Detection useful while making sure TR workouts don’t accidentally become harder than the workload my coach planned.
Thanks again for the clarification — both perspectives were very helpful.