Client Check-ins and Progress Reviews
For online coaches, weekly check-ins are the heartbeat of the service. They are also the single biggest writing burden — five clients means five thoughtful replies on the same day. AI lets you produce more personal, more specific check-in replies in a fraction of the time.
What You'll Learn
- A repeatable structure for high-quality check-in replies
- Prompts for handling stalled progress, big wins, and rough weeks
- Reviewing biometric and training-log data with AI
- Avoiding the "AI-generic" tone that clients can spot in seconds
The Anatomy of a Good Check-in Reply
Whether AI writes the first draft or not, a great check-in reply includes:
- Acknowledgment of how the week actually went, in their words
- One specific observation about their data or progress
- One priority adjustment for next week (training, nutrition, or recovery)
- One question that opens conversation (or one accountability note)
- A close that fits your coaching voice
If your check-in replies sound the same to every client every week, you're losing the relationship. The structure above keeps it personalized.
Prompt: Standard Check-in Reply
Use this prompt for the typical "decent week, on track" check-in.
You are my coaching assistant. Draft a check-in reply to my client. Use the structure: acknowledge → one observation → one priority for next week → one question. Tone: warm, accountable, plain English, no hype.
Constraints: under 200 words. Do not give medical advice. Do not invent results I haven't seen.
Client snapshot: [paste your saved client context block] This week's check-in: [paste their message] Their numbers (if relevant): [weight trend, sessions hit, protein average, sleep, mood 1-10]
Real Example
Client message:
Decent week. Hit all 4 sessions. Bench felt strong, hit 80kg x 5 (PR). Squats felt heavy though, only 100kg x 5 RPE 9 instead of the planned RPE 8. Sleep was bad Monday and Tuesday — work stress. Weight up 0.4kg. Stuck to nutrition mostly, missed protein on Wednesday.
Client snapshot you paste:
Tom, 31M, intermediate, 3 years training. Goal: lean gain (currently +200 kcal surplus). Knee pain in past, fully resolved. 4x/week upper/lower. Likes direct feedback, not hype.
AI draft:
Tom — solid week overall. Hitting all 4 sessions in a stressful work week is the win to underline. Bench at 80kg x5 is your PR, well done.
The squat read makes sense in context — two short-sleep nights early in the week show up in heavy lower work. RPE 9 instead of 8 isn't a failure, it's signal. Don't push through it on a 3-day fatigue debt; we plan with the body that turned up, not the one we wanted.
Weight up 0.4kg fits the surplus — good sign you're not fat-gaining too fast. Wednesday's protein miss is the only flag worth fixing. Easy lever: a 30g whey or skyr at the same point each day so it's not a decision.
Next week's priority: protein floor of 160g daily, no exceptions. Squat RPE stays at 8 — load doesn't matter, position does.
Question: how would you rate your work stress for next week vs this one — same, lower, higher? It changes how aggressive we get on the squat.
You'd polish the voice slightly, send it. Three minutes start to finish.
Handling a Rough Week
Bad weeks are where coaches earn their fee. AI helps draft the message, but the empathy must be yours.
Prompt
Draft a check-in reply for a client who had a hard week. Their message is below. Tone: empathetic, not pity. Acknowledge that the week was rough without amplifying the spiral. Reframe one thing constructively. Set a single, tiny, achievable target for next week — something they can absolutely do. Do not lecture. Do not give medical or psychological advice.
If the message contains language about mental health crisis or harm, stop and write only "[ESCALATE — message contains crisis indicators]" instead of a reply.
Client snapshot: [paste] Message: [paste]
That last instruction is important. AI can recognize crisis-language patterns. If a message contains anything alarming, you don't want a chipper reply — you want to flag it and respond personally with the relevant referrals.
Handling a Big Win
Don't let AI rob big moments of their realness.
Prompt
Draft a check-in reply for a client who hit a major milestone this week: [milestone, e.g., "first unassisted pull-up after 9 months of work"]. Tone: genuinely celebratory, specific to their journey. Reference how far they've come (use the timeline below). End with the next sensible target — one rung up the ladder, not three.
Client journey notes: [bullet points of their progress over the engagement]
For wins, the AI draft is a starting point. Add one personal detail AI couldn't know — the specific session they almost gave up, the in-joke between you, the photo from week 1 you'll attach.
Reviewing Training Logs and Biometrics
If your coaching app exports CSVs or your client uses a wearable that does, you can have AI review the data.
Prompt: Training Log Review
Below is a 4-week training log for a client. Identify:
- Sessions where load progressed
- Sessions where load stalled or regressed
- Patterns by day-of-week (e.g., always struggling on Mondays)
- Movements that may need a deload or substitution
Output: 5-bullet summary I can use as the basis of next week's plan.
[paste log]
Prompt: Wearable Data Review
Below is a client's weekly summary from [Whoop/Oura/Garmin]. Translate the data into 3 plain-language observations and 1 actionable suggestion. Avoid medical diagnosis. Frame in terms of training readiness, recovery, and consistency.
[paste data]
Avoiding the "AI-Generic" Tone
Clients can spot a robot reply. Common tells:
- Generic compliments ("Amazing work!" "You're crushing it!") with no specifics
- Bullet-list replies for emotional check-ins
- "Remember, every journey starts with a single step" cliches
- Identical structure week after week
Counter-moves:
- Reference their week verbatim — quote one phrase they used
- Vary opening lines — never start every reply the same
- Add one detail AI couldn't know — the in-joke, the recent context
- Keep your voice quirks — if you call them "kid" or use a specific phrase, AI usually drops these; add them back in
A useful rule: if the reply could go to any of your 25 clients with no edits, it's not personal enough.
Building a Check-in Workflow That Saves Hours
Here's how a coach with 20 weekly clients runs check-ins efficiently:
- Sunday evening: collect all check-in messages and biometric exports in a single doc, ordered by client
- Save your client context blocks in a Notion / Apple Notes folder with one file per client
- Batch through them: paste context + message + numbers, generate draft, edit, send
- Track patterns: at the end, AI reviews the full week of check-ins to flag clients who need a 1:1 call
Average time per check-in drops from 15-20 minutes to 4-6 minutes. For 20 clients, that's 4+ hours back per week.
Quick Reference Prompt Library
Save these in a notes app:
- "Draft a check-in reply for a stalled-progress week..."
- "Draft a reply that gently raises that the client missed 2 sessions..."
- "Summarize this client's last 4 weeks of check-ins into a 1-page progress report..."
- "Convert these check-in messages into a CSV table: week, weight, sessions, protein avg, sleep avg, mood..."
- "Rewrite this draft in a less hype-y, more grounded voice..."
Key Takeaways
- A great check-in reply has structure: acknowledge → observation → priority → question
- Always add at least one detail AI couldn't know — quotes, in-jokes, specific context
- Crisis-language indicators should escalate to you, not get auto-replied
- Wearable and training data can be reviewed by AI for patterns you'd miss
- Done right, check-in time drops by 60%+ without losing personalization