AI feedback tone
Every analysis comes with AI-written coaching feedback. The feedback tone setting lets you choose the voice that feedback is written in, so it sounds like your coaching — supportive, direct, technical, or entirely your own.
Each coach sets their own tone. Feedback for an athlete is written in the voice of that athlete’s coach. It changes only how the feedback reads — never the analysis itself, the numbers, or what appears on the athlete’s report.
Choosing a tone
Open Settings from your account menu (the avatar in the top-right) and find the Feedback tone section. Pick one of the presets:
- Balanced — calm, precise, and coach-like. This is the default, and it’s the voice your feedback already uses if you’ve never changed the setting.
- Supportive & encouraging — leads with what went well and frames misses as opportunities, while staying honest about the data.
- Direct & concise — short sentences, straight to the point: what worked, what didn’t, what to do next.
- Technical & analytical — backs every observation with numbers and explains the reasoning behind each recommendation.
When you select a preset, the exact instruction text that will be sent to the AI is shown below the selector — what you see is exactly what’s used. Save, and the new tone applies to feedback generated from then on.
Writing your own — “My tone”
If none of the presets sound like you, choose My tone and write your own instruction in the text box. A few tips:
- Describe the voice, not the content: “Warm but no fluff. Keep it under four sentences per section. Always end with one concrete next step.”
- Keep it short — up to 500 characters.
- Plain language works best. You don’t need prompt-engineering tricks.
Your text is saved to your own settings, so you can come back and edit it any time. If you switch to a preset later, your custom text is kept and will be there when you switch back.
The tone setting controls the voice the feedback is written in. What the feedback covers comes from the built-in review criteria for the session’s workout type (managed platform-wide as analysis profiles). To add your own context to those criteria for your own athletes, use analysis notes.