A new client tells you they can't do a push-up. You have two obvious options and both are bad: put push-ups in the program anyway and watch them do half-range reps on their knees for eight weeks, or replace push-ups with a machine chest press and never come back to it.
The third option is a ladder — a sequence of versions of the same movement, from one the client can do properly today to the one you actually want them doing, with a written condition for moving up each rung. This covers the ladder for each major pattern, how to judge readiness when you cannot watch the set, and what to do with a client who has not moved for six weeks.
A ladder progresses the movement; load progresses the rung
Two different things get called progression and they run on separate clocks.
Load progression happens inside a rung. The client stays on the goblet squat and the weight or the reps go up week to week, on whatever scheme you use — progressive overload for online clients covers that half, including double progression and how to read a stall in the numbers.
Movement progression is the rung itself: box squat to goblet squat to front squat. It is not scheduled and it does not run weekly. A client sits on a rung until they meet a standard you wrote down, which might be three weeks or might be five months.
Keeping the two separate matters because they fail differently. A client who is stalled on load usually needs rest or a smaller increment. A client who is stuck on a rung needs a missing quality trained — ankle range, overhead position, enough strength to hold their own bodyweight — and adding kilos to the rung below does not supply it.
Where regressing an exercise stops being coaching
Regressing an exercise because a client cannot yet perform it well is ordinary coaching. That is the whole job: find the version they can do properly, build it, move up.
Regressing an exercise because it hurts is not. Pain is not a technique fault you coach around with an easier variation, and swapping to a lighter version so the client can keep training through it is how a coach ends up managing a problem they are not qualified to manage. If a movement produces pain, the client sees a physiotherapist or a doctor, and you program around the area in the meantime on whatever advice comes back. Keep working with the practitioner rather than instead of them — how to program around client injuries goes through that handover and what stays inside your scope.
The line is worth stating plainly to clients at onboarding, because otherwise they report pain as "my knee felt weird" in a check-in note and expect a programming answer.
The seven ladders
These are the patterns worth carrying ladders for. Every exercise you prescribe sits somewhere in one of them, and once you have the seven written you stop inventing regressions on the spot at 9pm.
The gate column is what has to be true before the client moves up a rung. Each one is written so that the client can report it, because online they are the person watching.
| Pattern | Ladder, easiest rung first | Gate to move up |
|---|---|---|
| Squat | Box squat to a high box → box squat to a parallel box → bodyweight squat, full depth → goblet squat → front or back squat | 3 sets of 8 at the current rung, heels flat, knees tracking over the toes, and the last rep goes as deep as the first |
| Hinge | Hip hinge with a dowel on the back → 45-degree back extension or glute bridge → dumbbell Romanian deadlift → trap-bar deadlift from blocks → deadlift from the floor | The client feels the hamstrings, not the lower back, and can reach mid-shin without the back rounding |
| Horizontal push | Wall push-up → hands on a bench → hands on a low box → floor push-up → feet elevated, then weighted | 3 sets of 8 with the body in one line, chest to within a fist of the surface, hips not sagging or leading |
| Vertical push | Half-kneeling single-arm dumbbell press → seated dumbbell press → standing dumbbell press → barbell overhead press | Arms reach fully overhead without the ribs flaring or the lower back arching to get there |
| Horizontal pull | Inverted row on a high bar, body near vertical → bar lowered → feet elevated → single-arm dumbbell row → barbell row | 3 sets of 8 at the current bar height, chest touching, body straight from ear to heel |
| Vertical pull | Lat pulldown → band-assisted chin-up → 5-second eccentric chin-up → full chin-up → weighted chin-up | 3 controlled eccentrics of 5 seconds, or a pulldown set of 8 at bodyweight |
| Loaded carry | Suitcase carry, light, 20m → farmer's carry → front-rack carry → heavy or offset carry | 40m without the shoulder shrugging, the torso leaning away from the load, or the breath being held |
Two notes on reading that table. The rungs are versions of the same movement, not a warm-up sequence — a client on rung two trains rung two, not rungs one through four in the same session. And the top rung is the full expression of the pattern, not a goal every client needs. Plenty of clients finish a year of coaching on a goblet squat and a trap-bar deadlift and get everything they came for.
Lower body: depth and the back are the whole gate
For the squat, the rung is chosen by depth the client can control, not by the implement. A box is the most useful regression there is because it makes depth a fact rather than a judgement — you set the height, they sit to it, and the standard is unambiguous when you are not in the room. Lower the box before you add the dumbbell. A client squatting to a low box with bodyweight is closer to a front squat than a client goblet squatting 20kg to a quarter-depth.
For the hinge, the gate is the back position and where the client feels the work. The most common error is putting an untrained client straight onto a barbell deadlift from the floor, which asks for more hip and hamstring range than most people have and gets paid for in lumbar flexion. Start at a height they can reach, which is what the blocks are for, and lower the bar as the range arrives.
Both patterns share a trap: a client who cannot yet do the rung will do it anyway if you write it in the program, because they assume the prescription is correct and their body is the problem. Only prescribe the rung you have evidence for.
Pushing and pulling: change the angle, not the range
Push-ups and rows are the easy patterns to progress well, because the angle gives you a continuous dial. Hands at chest height, then a bench, then a low box, then the floor is four rungs of the same exercise with no equipment and no half-reps.
Half-reps are what happens when you skip the dial. A client doing floor push-ups to a six-inch depth is not doing an easier push-up, they are doing a different exercise that trains the top third of the range and builds the habit you will spend two months undoing. Knee push-ups have the same issue in a milder form: the trunk position that makes a push-up worth prescribing is the first thing to go. Elevate the hands instead.
Vertical pulling is the pattern where the ladder matters most, because the gap between a lat pulldown and a chin-up is enormous and nothing between them is obvious. The rung that does the work is the slow eccentric: the client jumps or steps to the top position and lowers for five seconds. Three of those, controlled, is a reliable signal that a full chin-up is close. Band assistance is useful but flatters the bottom of the range, where the band gives the most help and the client needs the most.
Judging readiness when you cannot watch the set
You need two pieces of evidence and both are cheap.
The first is the log. Write the gate into the prescription as a countable thing, so meeting it shows up as numbers rather than as a feeling: three sets of eight at this rung, or three five-second eccentrics. A client cannot misreport reps the way they can misreport how a set felt.
The second is one video, at the point the log says the gate is met — not weekly, and not on every exercise. Ask for the last working set, filmed side-on from about three metres, in one take. That is a small enough request that clients actually do it, and it answers the only question you have: should this person move up. How to run remote form checks covers getting usable footage and giving feedback the client can act on.
Then ask the client the right questions in the check-in. "How did the squats feel?" produces nothing. "Did your heels stay flat on every rep of set three?" produces an answer you can use, because it is a yes or a no about something they can notice while they are doing it. Two or three of those per rung is enough.
One rule underneath all of it: the client moves up when you say so. Never write "progress when it feels easy" into a program. A client's own sense of readiness is what put them on a barbell in week one of their last gym membership.
Most coaches progress technique too fast and load too slow
Both errors come from the same place — the ladder feels like an insult to the client, and the weight feels like the thing they are paying you for.
So the client gets moved from a box squat to a barbell in three weeks because a box squat looks like you are not coaching hard enough, and then sits at 30kg for four months because you cannot see them and the increments feel risky. That is backwards. A rung is a skill and skills take weeks: three to six weeks on a rung is normal and three months is not a failure. Load inside a rung, for a client in their first year, should move nearly every week, and 2.5kg on a goblet squat is not a gamble.
The honest test on any rung is whether the last rep looks like the first. When it does, the load is the thing to move. When it does not, no amount of extra weight fixes it and the client is on the wrong rung.
A client stuck on the same rung for six weeks
Six weeks with no movement is a real problem, but it is four different problems and they need different answers. Work through them in order.
- Are they training it? Count the sessions where that exercise was actually logged over the six weeks. Two sessions a fortnight is not a stalled progression.
- Is your gate too strict? "Perfect form" is not a gate. If you cannot say in one sentence what the client has to demonstrate, they cannot pass it and you have no way to tell that they have.
- Is a specific quality missing? A chin-up that will not come is usually a lat and biceps strength gap; a squat that will not deepen is usually ankle range. That is a programming answer — add the accessory that trains the missing piece and keep the rung where it is. Hypertrophy programming for clients covers building that supporting volume without wrecking the rest of the week.
- Is the next rung too big a step? Most six-week stalls are this. Split it: a push-up from a bench to a low box is one rung in most ladders and two in a real client's program.
If none of that moves it, the last answer is legitimate — this rung is the client's full expression of the pattern for now, and you stop treating it as a waiting room. A client with the ankle range they have may never squat to depth with a bar on their back, and a program built on goblet squats and split squats will still get them strong.
Write each ladder down once and reuse it
The reason coaches invent regressions on the spot is that the ladder only exists in their head, so every new client is a fresh decision at the end of a long day. Write the seven ladders once, with the gate for each rung, and the decision becomes a lookup.
In practice that means keeping every rung as its own exercise in your library, with the demonstration video and the written cue attached, so prescribing rung three is the same amount of work as prescribing rung five. In Fitsly's training tools you can also save alternate exercises against an exercise — the client can swap to one mid-session — and attach a coaching note the client reads on the set, which is where the gate belongs so they know what they are working towards. The same set of rungs then feeds every client's program, which is most of what writing programs faster for multiple clients comes down to.
Progressing a client with two light dumbbells
A client at home with a pair of 8kg dumbbells cannot progress by load for very long. The ladder is not a fallback for them; it is the main progression, and you run it by making the same movement harder at a fixed weight.
Five levers, roughly in the order they cost the least:
- Range. Deficit split squats off a step, or push-ups with the hands on books, add range before they add anything else.
- Leverage. Feet-elevated push-ups, incline instead of flat, a longer lever on any carry. The dumbbell has not changed and the exercise is meaningfully harder.
- Base of support. Two legs to a split stance to a Bulgarian split squat to a single-leg Romanian deadlift. This is where most home progression comes from, and it roughly doubles the load per limb without a heavier dumbbell.
- Tempo and pauses. A three-second lowering phase or a two-second pause at the bottom. Write the count in the prescription so it is not left to feel.
- Density. Same work, shorter rest, or more sets in the same session. Use this last — it changes what the session trains.
Move one lever at a time and prescribe it yourself. A client told to "make it harder" changes the tempo, the stance and the rest in one session, and the log after that tells you nothing.
