How to Write a Training Program: A Coach's Guide
A no-fluff framework for coaches: run a needs analysis, pick the split, set volume and progression, and deliver a program your client will actually follow.
Start with a needs analysis, not a template
The difference between a program that works and one that gets ignored is the ten minutes you spend before you write a single exercise. A needs analysis tells you what this specific client needs, what they can realistically do, and what will break if you get it wrong.
Cover four things: the goal, the schedule, the history, and the constraints. What does the client actually want in the next 12 weeks (strength, hypertrophy, fat loss, or general performance)? How many days a week can they train, and for how long? What is their training age and injury history? What equipment do they have, and what do they hate doing so much they'll quit? Write the answers down. Everything downstream depends on them.
Turn the goal into a training bias
Each goal changes the levers you pull. Strength means lower reps, higher intensity, longer rest, and heavy compound lifts. Hypertrophy means moderate reps, moderate-to-high volume, and shorter rest. Fat loss is mostly a nutrition and adherence problem, so the training keeps muscle and burns calories without wrecking recovery. Performance depends on the sport, but usually blends strength, power, and conditioning. Name the bias out loud so every later decision has a reference point.
Choose frequency and split by availability, not by ideals
The best split is the one your client will actually complete. Match frequency to the days they gave you in the needs analysis, then pick a structure that fits.
- 2 days: two full-body sessions. Simple, high return, hard to skip.
- 3 days: full-body x3, or push/pull/legs for slightly more advanced clients.
- 4 days: upper/lower x2. My default for most intermediate clients.
- 5-6 days: push/pull/legs run twice, or an upper/lower/full hybrid, for committed lifters with real recovery capacity.
Higher frequency lets you spread volume across the week so each muscle gets hit twice, which most people respond to better than once-a-week bodypart splits. But a perfect 5-day plan a client hits three times a week is worse than a 3-day plan they finish. Program for the calendar you have.
Select exercises compound-first, then individualize
Build each session around one or two big compound lifts that drive the most adaptation for the time spent: squat, deadlift, hinge, bench, overhead press, row, pull-up. These are the backbone. Then add accessories to fill in the muscles the compounds miss or under-stimulate.
Individualize from there. If a client has cranky shoulders, swap barbell bench for dumbbell or a neutral-grip press. No rack at home? Goblet squats and Romanian deadlifts with dumbbells cover most of the job. If someone genuinely hates back squats, front squats or a leg press keep the stimulus without the fight. The exercise is a tool to load a movement pattern, not a religion. Pick the variation the client can perform well, load safely, and progress.
Set volume, intensity, and sets/reps by goal
Volume (hard sets per muscle per week) is your primary dial. A practical range for most clients is 10-20 hard sets per muscle per week, starting at the low end and adding as they adapt. Intensity is how heavy relative to their max, usually managed by reps in reserve (RIR): leave 1-3 reps in the tank on working sets so quality stays high.
- Strength: 3-6 reps, 3-5 sets on main lifts, 1-2 RIR, rest 2-4 minutes.
- Hypertrophy: 6-12 reps (up to 15-20 on isolations), 3-4 sets, 1-2 RIR, rest 60-120 seconds.
- Fat loss: keep hypertrophy-style loading to preserve muscle; the deficit does the fat loss, not high-rep circuits.
- Performance: lower reps and full rest on power work, then goal-specific conditioning.
Don't max out volume in week one. Leave room to add sets over the block, because adding stimulus is your progression currency and you can only spend it if you saved some.
Pick a progression model and make it concrete
A program without a progression rule is just a workout. The client needs to know exactly how to get harder over time. The core principle is progressive overload - do a little more over time - and there are three clean ways to write it.
Linear progression
Add a small fixed amount each session while reps and sets stay the same: 2.5kg on upper-body lifts, 5kg on lower-body, week after week. Perfect for beginners because they adapt fast. It stalls once the jumps outpace recovery, which is your cue to switch models.
Double progression
My default for intermediates and accessories. Give a rep range, say 8-12. Keep the weight until the client hits the top of the range on all sets, then add load and drop back to the bottom. It self-regulates and removes the guesswork of loading isolation work.
RPE / autoregulation
Prescribe effort instead of a fixed weight: work up to a set at RPE 8 (2 reps in reserve). On a strong day they lift more, on a tired day less, but the effort stays constant. It's the most flexible model and the most forgiving of bad sleep and stress, but it asks more of the client's honesty and experience. Use it with people who can gauge effort.
Add periodization without overcomplicating it
Periodization just means organizing training into phases so you keep progressing and don't burn out. You don't need a spreadsheet with Greek terms. Run 3-4 week blocks where volume or intensity climbs, then a lighter deload week (cut sets roughly in half, or drop intensity) to let fatigue clear and adaptation catch up. Most clients need a deload every 4-8 weeks, or whenever performance and motivation both dip. That's 90% of the benefit for beginners and intermediates. Save advanced block periodization for advanced athletes.
Fill in warm-ups, accessories, and conditioning
Every session opens with a warm-up: a few minutes to raise temperature, then ramp-up sets on the first compound (empty bar, then progressively heavier singles or triples up to the working weight). Skip the 20-minute mobility routine unless there's a specific need.
Accessories target weak points and the muscles the main lifts under-hit: arms, rear delts, calves, core, and single-leg work. Two to four accessories per session is plenty. Conditioning is programmed on purpose, not thrown in as punishment: 1-3 sessions a week of low-intensity cardio for the fat-loss and performance clients, kept far enough from heavy leg days that it doesn't sabotage recovery.
Write it so the client actually understands it
A program the client can't read is a program they won't follow. Be explicit. For every exercise give sets, reps, the target RIR or RPE, and rest. Add one short cue where it matters ("brace before you descend"). Spell out the progression rule in one sentence at the top: "When you hit the top of every rep range, add weight next week." Link a video for anything unfamiliar. Assume the client will read it alone in a busy gym with no one to ask.
Use templates without making programs generic
Templates are how you stay efficient and consistent, and there's nothing wrong with them. The mistake is shipping the template unchanged. Keep a base structure per goal, then individualize the parts that matter: swap exercises for the client's equipment and injuries, tune volume to their recovery, and set starting loads from their actual numbers. The skeleton is reusable; the fit has to be bespoke. Building programs from a real exercise library inside coaching software makes this fast - you clone a proven template, swap a few movements, and deliver in minutes instead of rebuilding from a blank page every time.
Review and adjust from check-in data
The program is a hypothesis; check-ins tell you if it's true. Each week or two, look at what actually happened: Did they complete the sessions? Are loads and reps trending up? How's sleep, soreness, and motivation? Is bodyweight moving the right direction for the goal? If progress is happening and recovery is fine, leave it alone and let the progression run. If lifts stall for two-plus weeks despite good effort, change one variable, not five: add a set, swap a stalling exercise, or schedule a deload. Small, data-driven tweaks beat rewriting the whole plan on a hunch.
Deliver it: app vs PDF
A PDF is fine to start, but it's a dead document. The client can't log inside it, you can't see if they trained, and every edit means resending a file. A coaching app fixes all of that: the client sees today's session with videos, logs sets and reps as they go, and you get live completion and performance data that feeds your next adjustment. For online coaching especially, that feedback loop is the product. If you're moving from in-person to remote, our guide to starting online fitness coaching covers the delivery side in depth.
A worked example
Say your client is Maya: intermediate, wants hypertrophy, can train four days for 60 minutes, has a full gym, and has a slightly irritable left shoulder. Needs analysis done. Goal bias: hypertrophy. Frequency: four days, so upper/lower x2. Exercises: neutral-grip dumbbell press instead of flat barbell (shoulder-friendly), plus rows, RDLs, leg press, pulldowns, and targeted arm and rear-delt accessories. Volume: start at 12 hard sets per muscle per week, room to reach 18. Rep scheme: 6-12 on compounds, 10-15 on isolations, all at 1-2 RIR. Progression: double progression - hit the top of the range on every set, then add load. Periodization: build volume for three weeks, deload in week four. Delivery: assigned in-app with videos and logging so you can watch her numbers climb and pull the shoulder exercise if it flares. That's a complete, individualized program written in one pass from a clear framework.