How to do a monthly review of your personal principles
A weekly review answers a small, useful question: did I live this principle this week, and what was going on when I didn't. A yearly review answers a much bigger one: are these still the right principles to have at all. A monthly review sits between them and answers something neither of the other two can: is this getting better, staying the same, or getting worse. That question needs more than one week of evidence to see and more than eleven months to still be useful by the time you act on it.
What four separate weekly reviews cannot tell you
The weekly review works precisely because it treats every miss as its own incident: what was happening, not why do I keep failing. That design choice is correct at the weekly grain — it keeps the review honest instead of turning it into self-judgement. But it has a blind spot built into the same design. If you missed the same principle for the same reason four weeks running, a weekly review shows you four separate, reasonable-sounding exceptions. It never shows you the four next to each other, which is the only way to notice they are actually one thing.
A slipping standard and a bad week look identical from inside a single week. "I didn't protect my mornings because of a rough night's sleep" is a perfectly fair account the first time. It is still a fair account, word for word, the fourth time — and that repetition is the entire signal. A weekly review, doing its job correctly, will not surface that on its own. Something has to look at more than one week at once, and a year is far too long to wait for that look.
Gather four weeks of material, and read them together
Do not try to summarise the month from memory — a month reliably rounds off into "fine, I think" regardless of what actually happened, the same way a year does. If you have four weekly reviews or a decision journal, that is the raw material. Pull them up side by side and read for these specifically:
- Same principle, same excuse, more than once. One rough week is circumstance. The identical explanation twice is a condition worth naming and designing around, not a coincidence worth waiting out.
- A trend inside one principle's line, not just its average. Four scores of 6, 5, 4, 3 and four scores of 4, 5, 4, 5 can average out the same, but one is sliding and one is stable. You will only see the difference by reading the sequence, never from a single monthly number.
- Anything that produced no data for the whole month. A principle you failed is visible and gets attention on its own. A principle you quietly stopped engaging with produces nothing to notice — no low score, no exception to explain — which is exactly what makes it the more dangerous of the two. A month of silence is worth flagging on purpose, because nothing else will flag it for you.
- A behaviour standard you already rewrote mid-month. Check whether the rewrite actually helped or just moved the same problem one week down the road. A standard that needed two rewrites in a month is telling you something about its design, not about your effort.
Pick one change, not a list of them
The same rule that keeps a weekly review from turning into a to-do list applies here at a slightly larger scale: end with exactly one adjustment. Rewrite the standard that keeps producing the same excuse. Retire a principle that has not been tested all month because the situation that would test it no longer exists. Or, if nothing needs changing, say so and stop — a month with nothing to fix is a legitimate outcome, not a sign you looked too quickly.
Do you actually need this third layer
Not everyone does, and it is worth being honest about that before adding it. If your weekly reviews are already producing real edits and nothing feels like it is quietly sliding, a monthly pass is one more ritual competing for the same ten minutes, and the extra layer is not paying for itself. Add it specifically when you notice the opposite: weekly reviews that read like a string of separate, reasonable incident reports, after which you genuinely cannot tell whether the month as a whole was an improvement or a decline. That inability to tell is the actual symptom a monthly review treats — not a fixed schedule you owe the practice out of thoroughness.
The same failure mode that kills a weekly review can kill this one too: turning it into a scorecard. The point of putting four weeks side by side is to notice a condition worth changing, not to compute a monthly grade. If the exercise starts feeling like a performance review of yourself, you have rebuilt the exact problem the weekly format was designed to avoid, just on a longer cycle.
When to do it
Pick a fixed point that follows your weekly rhythm rather than fighting it — the first weekend after the fourth weekly review of the month, for instance, so there is always a full set of material waiting rather than three reviews and a gap. Keep it to fifteen or twenty minutes. You are reading and comparing material that already exists, not writing anything from scratch, and a review that requires a fresh hour is a review that stops happening by the second month.
This is close to what CreedOS keeps visible without being asked: the review page shows the week-by-week trend for each creed alongside the average, so a slide across four weeks is something you see rather than something you have to reconstruct from memory, and an issue queue quietly surfaces the creeds that have gone idle for a while rather than waiting for you to notice the silence yourself. None of that replaces actually sitting down once a month and reading — it just means the material is dated and in your own words when you do. It's free, with no in-app purchases.