Most people meet the phrase as an instruction to add weight. Put a little more on the bar than last time, every week, forever, and you will get stronger forever. It works for a few months. Then a week arrives where the bar does not move, and the instruction has nothing to say about it.
The principle underneath is broader than that, and knowing the broader version is the difference between a lifter who has one route up and a lifter who has five.
The demand has to rise, and weight is one way to raise it
A muscle changes because it is asked to do something it is not already comfortable doing. Keep asking for exactly what it can already handle and it has no reason to become anything else. That is the whole of the principle: the demand rises, and the body follows.
The weight on the bar raises the demand. So does an extra rep at the same weight. So does an extra set, a longer range of motion, a slower lowering, a shorter rest, and another session in the week. Every one of those is more work than you did last time, which is what the principle is asking for.
That is not a technicality. It has been tested. 43 trained lifters, each with at least a year of lower body training behind them, were split for eight weeks: one group added weight and held the rep range, the other added reps and held the weight. Strength went up around twenty kilograms in both, with a small point estimate in favor of adding load, and the muscle thickness differences between the groups came out under a millimeter almost everywhere. The authors' own conclusion is that both are viable1.
Then the same comparison was run again on 39 previously untrained men and women, over ten weeks, with each person progressing load on one leg and reps on the other. Both legs got stronger. Both legs got bigger. Neither beat the other2.
Progressive overload is not a weight going up. It is a demand going up, and there are more ways to raise it than there are plates on the rack.
Load still has to move eventually
That is the part it would be dishonest to leave out. Reps and load are interchangeable enough over an eight or ten week study. They are not interchangeable over a training life. If your goal is a heavier one repetition maximum, the load has to rise at some point, because lifting a heavy weight is a skill you practice by lifting heavy weights4.
So the honest version is a sequence rather than a choice. Reps buy you the control to hold a heavier weight in a position you can still keep. Load turns that control into strength. The American College of Sports Medicine puts the same idea as a rule of thumb: raise the load once you can carry the current one one or two reps past the target, and not before3.
What to do with this in your own training
- Pick one variable per lift and stay with it. Trying to add weight, reps, sets and range of motion in the same week means you cannot tell which one moved you, and the week after that has nothing left to give.
- Climb reps before you climb load. It is the cheaper of the two and it leaves the position intact. When the range is finished on every set, the load has been earned.
- Take the smallest step your gym can make. The standing recommendation is a percentage3, and a percentage of a barbell is usually a number nobody can load. The smallest pair of plates on the rack is a step you can actually take, and small steps mean more weeks where the reps are good.
- Let a lift that will not move tell you something. One flat week is noise. A month of flat weeks is the lift, the recovery or the effort reporting, and none of those is fixed by adding weight anyway.
You will not progress every week, and that is not the plan failing
This is the part the phrase does the most damage with. Read as an instruction, it says every week must be bigger than the last, so a week that is not bigger is a week you failed. Anybody who has trained for a while knows that is not how it goes, and the coaching advice says so plainly: do not force progress every week, and do not chase the number for its own sake4.
Overload also has a partner, and the phrase does not name it. The joint consensus of the European College of Sport Science and the American College of Sports Medicine opens on exactly that point: training has to involve overload, and it has to avoid excessive overload combined with inadequate recovery5. A short dip in performance that clears with rest is an ordinary part of training. One that does not clear is a different thing wearing the same clothes5, and the first thing either takes from you is the quality of the sets you were relying on.
The question is never whether this week was heavier. It is whether this month asked more of you than last month did, and whether you turned up for enough of it to find out.
What Reckoner does with it
Reckoner reads what you actually lifted, set by set, and works out each lift's next step from that rather than from the date. The rule it used is printed beside the decision, so you can read it and disagree with it. Which variable moves, in what order, and why reps come first is the whole subject of its own guide.