Progressive overload is the least complicated idea in strength training and possibly in all of human endeavour: do slightly more, over time, than you did before. More weight, more reps, more sets, better control, shorter rest. Pick one. Move it. Repeat until strong.
The complication is entirely clerical. To do slightly more than last time you must know what last time was: every exercise, every session, far enough back to see a direction. And that is a trivial amount of arithmetic bolted to an impossible amount of memory. Memory is where it dies. Memory is where most things die.
The spreadsheet, and its six-week lifespan
Nearly everyone tries a spreadsheet. Nearly everyone abandons it around week six, and the reason is always the same, and it is never discipline: the spreadsheet is not open when you need it.
You need last week's numbers at the precise moment you're standing at the rack, chalk on your hands, breathing hard, ninety seconds until the next set. Anything that requires unlocking a phone, finding a browser tab, scrolling to the right row and reading across a column does not survive contact with an actual training session. It survives six weeks on enthusiasm, because enthusiasm is a fuel, and then the tank is empty and the spreadsheet joins its ancestors.
The fix is not becoming a better person. Better people have the same problem. The fix is that the previous numbers must already be there, sitting beside the set you're about to do, with no navigation between you and them. That is the entire secret, and everything below is footnotes to it.
What actually needs tracking
Less than you think. Considerably less than the spreadsheet thought.
Per set: weight and reps. That's the whole ledger. Everything else is derived.
Optionally, RPE: one number for how hard it felt. The single highest-value optional field, because it separates "3×8 at 80 kg that flew up" from "3×8 at 80 kg that nearly ended me." Identical rows in any log without it. Very different afternoons.
Not warm-ups. Log the working sets. Warm-ups say nothing about progression, and logging them turns a twenty-second job into a two-minute one, which (see above) is how logging dies.
Not a hand-maintained estimated 1RM. It's a fine summary line on a chart. Chasing it week to week by hand is how a hypertrophy block quietly becomes a series of near-maximal singles and a series of near-maximal singles quietly becomes physiotherapy.
The three views that do all the work
1. Last session, inline. Previous weight and reps beside the set you're about to perform. This is the one that makes progression automatic: you don't decide anything, you see 3×8 at 80 and put 82.5 on the bar, and the decision was made by a screen while you were thinking about nothing at all. That is the ideal amount of thinking to do between sets.
2. Exercise history. Every instance of the lift, in order. This is where you discover you've done 80 kg for three straight weeks while sincerely believing you'd been adding weight. The log doesn't share your optimism. That's what it's for.
3. The twelve-week trend. Not the session, not the week. Individual sessions are ruled by sleep, stress, caffeine and whether you remembered lunch. Twelve weeks is long enough for the line to be signal, and the line either goes up or it doesn't. Lines are refreshingly incapable of making excuses.
Why twelve weeks, spelled out
A bad session tells you nearly nothing. A good one tells you slightly less, because you'll over-read it, and everyone over-reads the good ones; it's the most human error in the whole enterprise.
Training data is noisy enough that short-window comparisons actively mislead: check your progress every session and pure variance will tell you you're regressing about a third of the time, and you'll rewrite your program in response to nothing, and then you'll never learn whether the old program worked, because you shot it before it testified.
Where the food comes in
The one thing a training log alone can never tell you: whether a stalled lift is a programming problem or an eating problem.
A plateau in a deficit is expected and usually fine. The identical plateau at maintenance means something else entirely. Same observation, opposite prescriptions. And the training log, by itself, cannot tell them apart, which is the whole argument for keeping training and food in one app.
Doing it in OneRep
Open a routine and start. Previous weight and reps sit beside every set, so the progression decision is made by what's on the screen. The rest timer runs itself. History and volume trends are a tap away, and the progress view puts strength, consistency and body metrics on the same twelve-week timeline as your food.
It logs offline, too, which sounds minor until you meet your first basement gym, and then it is the only feature that exists.
Try it: Free, no trial window.
Compare: OneRep vs Strong · OneRep vs Hevy