The weekly numbers tell you what is happening. The monthly review is where you decide what to do about it.
Without a fixed slot, this becomes something that happens when there is time, which means when there is a problem, which is too late to be a review and is really just a reaction.
Thirty minutes, same date every month, with an agenda.
The agenda
Minutes 1 to 5: the month's numbers. Revenue, expenses, profit, cash position at month end. Written next to the same figures from last month and the same month last year if you have it.
Do not analyze yet. Just look at them.
Minutes 5 to 12: what moved and why. Take the two largest changes, up or down, and write a one-sentence explanation for each.
The discipline is in being specific. Revenue was down is not an explanation. Revenue was down because two projects slipped into next month and one client did not renew is. The first tells you nothing; the second tells you whether it is timing or attrition, which are completely different problems.
Minutes 12 to 20: the assumption check. Take one assumption from your blueprint and compare it with what actually happened. Price, conversion rate, cost to deliver, customer lifetime, cost to acquire — rotate through them.
This is the most valuable segment and the one most likely to get cut. It is where the business's model gets corrected against reality rather than defended.
Minutes 20 to 27: one decision. Based on the above, what is the one thing you will change or start this month?
One. A review producing eight action items produces zero. Write the decision and how you will know in thirty days whether it worked.
Minutes 27 to 30: the calendar and compliance check. Filings due, renewals, insurance, tax deadlines, subscriptions worth cancelling. Two minutes that prevent the expensive surprises.
Write it down
Same document, one page per month, appended over time.
The compounding value is in the sequence. Reading twelve of these consecutively shows you patterns invisible at any single point — seasonality, recurring excuses, decisions made repeatedly and never followed through.
It also shows you the quality of your own reasoning. Reading a decision you made eight months ago, next to what you predicted would happen, is a genuinely uncomfortable and genuinely useful education in your own judgment.
If you have partners
Do it together and keep the record. It converts the ambient anxiety of how are we doing into a scheduled, evidence-based conversation, and it removes an enormous amount of the friction that builds up when two people are separately guessing.
Protect the slot
The reason to fix the date is that this will always lose to whatever is urgent. It is a small, boring block that produces one decision a month made on evidence rather than mood — which, over a year, is twelve decisions better than the alternative.