Your monthly business review is not a meeting. It is a reality check.

What to know

Imagine a fictional repair shop whose monthly review shows sales roughly on plan, fewer repeat bookings, and more unpaid invoices. The figures do not prove that service quality has slipped or customers are leaving; they identify questions worth checking. The owner decides to contact recent customers, assigns the calls to the service lead, and sets a date two weeks later to review what was heard and whether repeat bookings changed. That is more useful than adding “improve retention” to a general task list: it names a test, a person, and a time to learn. In your own review, separate observed signals from explanations, then choose one action that is affordable to reverse if the explanation proves wrong.

Put it into practice

Start with a question, not the spreadsheet

Before opening reports, write down one decision the business needs to make this month. Keep it narrow enough that the review could change what you do. “Should we keep offering same-day delivery?” gives you a choice to examine; “How is the business doing?” can lead anywhere.

Try this now: name the decision you are avoiding and the signal that might help you make it. If you do not track that signal, improving the tracking may be the sensible next step. A missing measure is not a reason to pretend you know the answer.

Choose a few signals that connect to the question

As a rule of thumb, more numbers do not necessarily make a review more useful. Choose a small set that could illuminate the decision: perhaps sales, cash collected, overdue invoices, repeat orders, delivery delays, or hours spent on a service. Compare each with a relevant reference, such as your plan or the previous month. A year-over-year comparison may help when periods are genuinely comparable.

For this month, choose up to three signals and note the comparison you will use. Flag anything that could distort the picture, such as a seasonal rush, a one-off purchase, or a shortened workweek. After the review, ask whether each measure helped with the opening question. Drop measures that add noise; add one only to fill a real gap.

Separate what happened from what you think it means

A changed number is an observation; its cause is a possible explanation. Keeping those separate can help reduce the risk of treating a hunch as a conclusion. Fewer repeat orders, for example, might reflect dissatisfaction, seasonality, a longer interval between purchases, or incomplete records. The number alone does not identify the cause.

Write “Observed” and “Possible explanation” in separate columns. Then ask what small check could distinguish between explanations. In a fictional independent café, the owner might notice fewer afternoon orders. Before changing opening hours, they could compare weekday and weekend patterns and ask a few regulars about their routines. This will not settle every uncertainty, but it can reveal whether the first explanation deserves more weight. Look for information that could change your mind, not just support your first guess.

Make one decision reversible where possible

End with one concrete action, a named owner, and a date to review what happened. “Improve customer retention” is a broad intention. “The service lead will contact ten recent customers by Friday and summarize recurring feedback at next month’s review” defines a manageable check. Keep its scope affordable in time and money, especially when the cause is uncertain.

If a change is costly or hard to undo, do not force a verdict from a thin signal. Gather more information, set a spending limit, or try a smaller version first. For a low-risk choice, state what you expect to see and what would make you stop. That makes it easier to reassess rather than letting a temporary adjustment become permanent by default.

Return to the decision, not just the task

At the next review, check whether the action happened, what it revealed, and whether the original decision still makes sense. If it did not happen, identify the obstacle; if the result is unclear, decide whether to extend, change, or stop the test. Record why, so the issue does not return as a vague worry. The aim is not to fix everything at once, but to leave with a decision and a useful next check.