The weekly business review checklist for consumer brands: eight items and what a red flag looks like

Meet Sam. He runs a made-up hand soap brand.

Monday, 9 a.m. The weekly meeting opens with good news. Sales hit $61,000, a record week.

Then his ops lead asks where that number came from.

Amazon says $38,000. The store says $19,000. That adds to $57,000.

Nobody can say where the other $4,000 came from.

Forty minutes later, the meeting ends. Nothing was decided.

You have probably sat in that meeting. A good weekly review fixes it in under an hour.

A weekly business review is a short meeting where a brand checks the same numbers every week, on the same basis, and writes down what it decided.

This post is the checklist. Eight items, in order. For each one: what to look at, the question it answers, and what a red flag looks like. Print it, or copy it into your meeting doc.

Before the meeting: three ground rules

These three rules matter more than any single item.

• Use closed weeks only. A closed week is one that ended at least a day ago, Monday to Sunday. Same-day data is still coming in. Comparing it to a full week makes things look worse than they are.
• One basis per number. A basis is what kind of number it is. Orders placed, cash paid out, or gross sales before refunds. Pick one for each line, write it down, and never mix them.
• Build the numbers before the meeting. If the review is spent making the numbers, nobody has time to decide anything.

The checklist on one page

Item 0 comes first on purpose. Everything after it is only as good as the data. It is also the item most teams skip.

0. Data health

Look at: when each source last updated. Amazon, the store, each ad account, email, the books.

The question: is every number in this review fresh?

Red flags:

• An ad account that has not updated in days. Often a login expired.
• A channel showing zero sales. Zero can mean no sales. It can also mean the sync broke. Treat it as broken until you know.
• A number that is exactly the same as last week. Real numbers almost never repeat.

1. Sales on one basis

Look at: net sales for the closed week, by channel, next to last week and the same week last year. Net sales means sales after discounts and refunds.

The question: how much did we sell, and where?

Red flags:

• The channels do not add up to the total. Something is double counted or missing.
• Amazon orders mixed with Amazon payouts. They are different numbers on different dates. Our guide to settlements vs orders shows why they never match week to week.
• A big jump with no event behind it. No promo, no launch, no press. Check the data before you celebrate.

All examples in this post use Sam’s made-up brand, with illustrative numbers. It sells hand soap on Amazon and its own store, and runs Meta and Google ads.

Example, illustrative: back to Sam’s $61,000 week. Amazon and the store add to $57,000. The missing $4,000 turns out to be wholesale orders, counted in the total but not shown as a channel. Small, but now everyone knows.

2. Profit after ads

Look at: contribution after ad spend, by channel. That is sales minus product cost, fees, shipping and ads. Our contribution margin guide walks through the math.

The question: what did we keep from what we sold?

Red flags:

• Sales up, profit flat or down. Growth is being bought, maybe at a loss.
• A product cost that has not changed in a year. Costs move. An old cost makes margin look better than it is.
• A channel with no cost data. A blank is not zero. Say what share of cost is missing instead of hiding it.

3. Ad payback

Look at: spend, new customers and the cost per new customer, by channel. If you can, look at payback days too. That is how long until a new buyer’s profit covers what it cost to get them.

The question: is ad money coming back, and how fast?

Red flags:

• Spend up 20% or more, new customers flat. You are paying more for the same buyers.
• Payback days getting longer, week after week. Slow drift is easy to miss.
• Each channel judged only by its own report. Ad platforms grade their own work, and they all claim the same sale. Meta cannot see an Amazon checkout at all. See incrementality vs attribution for a fairer read.
• Judging ads at day 7 when buyers take 60 days to pay back. Our post on first-order profit and payback explains the trap.

4. Stock

Look at: units on hand and on the way for each top product. Divide by units sold per day to get days of stock left.

The question: what runs out, and when?

Red flags:

• Days of stock shorter than reorder lead time. If the soap refill has 30 days left and takes 45 days to make and ship, it will run out. The only question is for how long.
• Stock on the way with no confirmed date. “Shipped soon” is not a date.
• On Amazon, a top product near zero. Amazon says an ad will not display if the product is out of stock (Amazon Ads FAQ). Amazon Ads bulksheets now show which advertised products are not eligible, and why (Amazon Ads). Organic rank can slide too, and rank is slow to win back.
• On Meta and Google, ads still sending people to an item that is sold out on your store. A regular ad keeps spending until someone stops it. In Meta, an automated rule can turn an ad set off when it meets conditions you set (Meta Business Help).

5. Repeat buying

Look at: new and returning customers by channel. Then the share of each month’s new buyers who came back within 30, 60 and 90 days.

The question: are the people we paid to get coming back?

Red flags:

• Newer buyer groups reorder less than older ones did at the same age. Compare day 60 to day 60, never a young group to an old one.
• Subscription cancels up after a price change.
• Returning customers falling while new customers rise. Fine for a week. A problem for a quarter.
• On Amazon, missing buyer data. Amazon shares little about who bought. A low repeat rate there can be a data gap, not a real drop.

6. Cash

Look at: cash in the bank, payouts on the way, bills due, and cash tied up in stock. Payouts lag sales. Amazon settles seller accounts about every two weeks (Amazon’s guide to seller payments). Shopify counts payout time in business days, so weekends do not count (Shopify Help Center).

The question: how many weeks can we run at this pace?

Red flags:

• Cash falling while sales rise. Growth is eating cash, often through stock and ad spend paid before the sales come in.
• A payout smaller than expected. Amazon holds part of your balance back as a reserve for refunds and chargebacks. Shopify payouts can drop for refunds, chargebacks, fees or a reserve (Shopify Help Center). Find out which one it was.
• A big stock order due before the cash from the last one is back.

Six items in, you have the numbers. Now comes the part that turns them into decisions.

7. What moved and why

Look at: the three biggest changes from items 1 to 6.

The question: for each one, what caused it, and how do we know?

Red flags:

• A move with no named cause. “Seasonality” is not a cause unless last year shows the same move.
• A cause nobody checked. “Meta drove it” needs a number behind it.
• Two causes averaged into a story. If Meta says one thing and Amazon says another, show both and say they disagree.

Example, illustrative: the soap brand’s store margin fell three points. The first guess was higher ad costs. The real cause was a shipping rate change on heavy orders, worth about $900 a week. Neither the ad account nor the store’s sales page showed it. It only came out when the shipping bills were read next to the orders. Nobody missed it. The data hid it.

8. Decisions

Look at: this week’s decisions, and last week’s.

The question: what did we decide, who owns it, and did last week’s decision work?

Write each decision down with four things.

1. What we decided. “Pause the two Meta ad sets losing money after fees.”
2. Why. The number behind it.
3. Who owns it.
4. When we check it, and what result means it worked. “Next Monday. Profit after ads on Meta should rise by at least $500.”

Red flags:

• A decision with no date to check it. It will never be checked.
• Last week’s decisions not reviewed. Then the team never learns which calls were good.
• The same problem decided three weeks in a row. Nothing is being done.

A 45-minute agenda

If any item runs long, park it. Give it an owner and look at it after the meeting.

Where Synthesis fits

Synthesis builds the numbers for this review. It connects a brand’s stores, ads, email and books. It checks each number against its source, such as Amazon sales against Amazon settlements. Its free weekly brief reads every channel on one basis, says what moved and why, and names one thing worth doing.

It does not run the meeting or make the calls. People do that. And some items, like cash in the bank, depend on which systems you connect.

Wrap-up

Same eight items. Same basis. Every week.

Start with data health, so you trust what follows. End with decisions, so the meeting changes something. Check last week’s decisions before you make new ones.

Print the one-page table. Use it this Monday.