The first question is what the numbers actually say. Revenue, enquiries, conversion rate, and the one measure you decided mattered when the quarter started. Read them against what you expected rather than in isolation, because a figure without a comparison carries no information.
The second is what you assumed that turned out to be wrong. Every plan rests on beliefs that have not been tested, and a quarter is long enough to have tested several. Somebody would pay that price. That channel would produce enquiries. A job would take four hours. Naming the assumption that failed is more useful than naming the outcome, because assumptions can be corrected and outcomes cannot.
The third is what you will stop. This is the question most reviews skip and the one that makes the others matter. Adding without removing produces a quarter that looks identical to the last one with more crammed in. If nothing came off the list, nothing meaningful was decided.
The fourth is what changes next quarter, stated specifically enough that somebody could check. Not improve marketing but publish one article a week. Not get more reviews but ask every customer at completion. A change nobody could verify was not a decision.
Do it on a fixed date rather than when it feels timely, because it will never feel timely. The first working week after each quarter ends is the usual choice, and putting the four dates in the calendar for the whole year at once is what makes it happen.
Look at the things that do not appear in the accounts. Which customers were difficult and why. What you found yourself avoiding. What broke and got patched rather than fixed. Whether you worked hours you would not want to repeat. Those shape the next quarter as much as the revenue figure does.
Write it somewhere you will find it, and keep the previous ones. The value compounds, because reading four consecutive reviews shows patterns invisible in any single one. The same problem appearing three quarters running is a different thing from a bad quarter, and only the record reveals it.
Keep it short enough that you will do it again. A page is enough. An elaborate template gets completed once and abandoned, and the discipline matters considerably more than the format.
Include one thing that went well and be specific about why, not for morale but because understanding what worked is how it gets repeated. A quarter where enquiries doubled is useful only if you can name what caused it.
Then act on it within the following week, while the conclusions are still fresh. A review conducted carefully and then filed is the most common failure mode, and it converts a useful habit into a ritual.
Involve somebody else if there is anybody to involve, even briefly. Explaining the quarter out loud surfaces things that reading the numbers alone does not, and a person who was not in the detail asks the questions you have stopped asking.
Keep the four dates in the calendar for the whole year rather than scheduling the next one each time. A review that has to be arranged is a review that gets postponed, and the postponement is always for a good reason.
Compare against the review before last rather than only the most recent one, since a two quarter gap makes drift visible that a single comparison hides.