How to Run a Monthly Review That Actually Improves Things
A monthly review that produces no decision is a reporting exercise, not a review. Thirty minutes, six things to look at, one decision recorded. That is enough to compound improvement over a year.
Why Most Reviews Fail
This guide is for supplement, Ayurvedic and herbal brand owners in India.
Monthly reviews commonly fail in one of two ways. Either they do not happen, because nothing forces them, or they happen and produce a discussion of numbers without a decision, which means nothing changes and the next review discusses the same numbers.
A review that works has three properties: it is short enough to happen, it looks at a fixed small set of things so the comparison is meaningful, and it ends with one recorded decision whose effect is checked at the next review.
One improvement per month compounds substantially over a year. Twelve discussions do not.
The Six Things to Look At
Fixed, so month-to-month comparison means something.
- Reorder rate by SKU, compared with previous months.
- Contribution per order after all variable costs.
- Complaints by cause category, compared with previous months.
- Returns by SKU and by route.
- Stock ageing, flagging anything approaching expiry.
- Acquisition cost and reorder rate by channel together.
The Records Check
Five minutes, and it prevents most of the problems that otherwise accumulate.
| Check | Why |
|---|---|
| Certificates of analysis received and read for every batch | Unread certificates defeat the purpose of having them |
| Batch log updated with key values | Drift is visible in a log and invisible in individual batches |
| Any licence or certificate approaching renewal | Expired documents in an otherwise good file undermine everything |
| Any change notified by the manufacturer | A change recorded in email and nowhere else is lost |
| Complaint log entries have batch numbers | Without them the entries cannot be used |
| Anything approaching expiry in stock | Decisions made early cost less than write-offs |
This part is unglamorous and is what distinguishes a brand that can answer a question in an hour from one that cannot.
Making a Decision
One per review, chosen deliberately.
Pick the Largest Actionable Thing
Not the most interesting. The largest complaint category or the weakest reorder rate usually points to it.
Make It Specific
A decision with an owner and a date, not an intention. “Look into packaging” is not a decision.
Record It
Written down where the next review will see it. Unrecorded decisions are not revisited.
Check It First Next Time
Begin the next review by asking whether last month’s decision was done and whether it worked.
Checking the previous decision first is what converts a review series into improvement. Without it, decisions are made and quietly abandoned.
What Not to Include
Three things that make reviews longer and less useful.
Revenue as the headline number, which is the least informative figure available and crowds out reorder rate. Report it, but do not lead with it.
Metrics you will not act on. If a number has never changed a decision, it does not need reviewing monthly.
Long discussion of a single month’s movement. Monthly data is noisy, and reacting to one month’s variation produces churn rather than improvement. Look at the trend across three or four months before acting on anything other than a clear problem.
Checklist
The review itself.
- Schedule it as a recurring commitment.
- Start by checking last month’s decision: done, and did it work?
- Review reorder rate by SKU against previous months.
- Review contribution per order.
- Review complaints by cause category.
- Review returns by SKU and route.
- Review stock ageing and flag anything approaching expiry.
- Review acquisition cost and reorder rate by channel together.
- Run the five-minute records check.
- Identify the largest actionable issue.
- Make one decision with an owner and a date.
- Record it where the next review will see it.
- Keep the whole thing to thirty minutes.
Found Something Worth Fixing?
Share what the data is telling you and our team will suggest specification or packaging changes worth testing.
Frequently Asked Questions
Why do monthly reviews fail?
Either they do not happen because nothing forces them, or they produce a discussion of numbers without a decision, so nothing changes and the next review discusses the same numbers. One improvement a month compounds; twelve discussions do not.
What should I look at?
A fixed set of six so comparison is meaningful: reorder rate by SKU, contribution per order, complaints by cause, returns by SKU and route, stock ageing, and acquisition cost with reorder rate by channel together.
What records check is worth five minutes?
Certificates of analysis received and read for every batch, batch log updated, licences approaching renewal, any manufacturer change notified, complaint entries having batch numbers, and stock approaching expiry.
How do I make the review produce improvement?
End with one decision that has an owner and a date, record it where the next review will see it, and begin the next review by checking whether it was done and whether it worked.
What should I leave out?
Revenue as the headline number, metrics that have never changed a decision, and long discussion of a single month’s movement. Monthly data is noisy, so look at a three or four month trend before acting on anything but a clear problem.