Most Amazon annual plans start with a number someone likes. "We did $4 million, so let's do $6 million." Then the team works backward, spreads the target across months, and calls it a forecast.
That is a wish with a spreadsheet around it. It ignores which products make money, how much inventory your cash can carry, and what peak season will actually do to your margins. By March the plan is stale, and every decision gets made on instinct again.
Q4 is the right time to fix this. You have a full year of data, fresh peak-season results, and enough time before January to place the orders that shape 2027. Here is a framework that starts from profit and cash, and ends with a plan that tells you what to scale, fix, and cut.
Start With Last Year's Contribution Margin by SKU
Revenue by product tells you what sold. Contribution margin tells you what earned. Pull twelve months of data at the SKU level and calculate, for each one: sales, Amazon fees, FBA fulfillment, cost of goods, returns, ad spend, and storage or aged inventory charges. What is left is the money that product contributed to your business.
If you have not built this view before, our guide on why contribution margin, not revenue, should drive every Amazon decision walks through the calculation line by line.
Then sort the list. You will usually see three groups:
- A small set of SKUs producing most of your contribution dollars.
- A middle group that earns a little and consumes a lot of attention.
- A tail that looks fine on the revenue report and loses money after ads, returns, and storage.
A revenue target tells you how big to dream. Contribution margin by SKU tells you where to spend.
Do this before you open a single forecast tab. Every later step depends on it.
Set the Ceiling Before You Set the Goal
Growth on Amazon is capped by two things that rarely appear in a revenue target: inventory and cash.
Inventory capacity
Estimate how many units you can realistically produce, ship, and store. Include supplier lead times, minimum order quantities, your FBA storage limits, and the weeks of cover you need to avoid stockouts. A stockout erases the rank you paid to build, which is why inventory planning that protects rank from a stockout belongs inside the plan, not next to it.
Cash
Inventory is cash sitting in a warehouse. Map the cycle: when you pay the supplier, when freight lands, when units check in, and when Amazon disburses the sales. Then set a firm number for the most cash you are willing to have tied up at any point in the year. That number is your ceiling.
Whatever revenue you can support under that ceiling is your realistic upper bound. Anything above it needs outside financing, and that is a separate decision with its own risks.
Sort Every SKU Into Scale, Fix, or Cut
Now combine the two views. Each SKU gets one label.
Scale
These products have strong contribution margin, healthy conversion, and room to grow in search. They get first claim on inventory, cash, and ad budget. Check that margin holds as spend rises. Our piece on scaling PPC without letting ACoS run away covers how to add budget while the math still works.
Fix
These products have demand but weak economics. The cause is usually specific: a high return rate, a bloated ad structure, an oversized package that inflates fees, or a listing that converts below its category. Assign each one a named problem, an owner, and a deadline. A fix with no deadline is a hope.
Cut
These products lose money after full costs and have no clear path to improve. Plan to sell through the remaining stock, stop reordering, and redirect the cash. Cutting is not failure. It frees capital for SKUs that can use it.
Be strict about the Fix label. If a product has been "almost there" for two quarters, it belongs in Cut.
Build the Budget From the Labels
With priorities set, the budget nearly writes itself.
Ad spend. Set a target ACoS for each product from its margin, not from a blended account average. The method is in the right way to set a target ACoS for each product. Scale SKUs get headroom. Fix SKUs get a capped test budget. Cut SKUs get only what clears inventory.
Inventory. Reorder quantities follow the Scale list first. Fix SKUs get smaller, more frequent orders until the problem is solved.
Creative and listing work. Spend where conversion gaps sit on products that already earn. Improving a listing for a SKU you plan to cut is wasted effort.
Launches. Treat new products as a separate line with their own cash allocation. Do not let a launch quietly drain the budget of your proven sellers.
Add a reserve of ten to fifteen percent of your annual ad and inventory budget that you do not assign in advance. Prices move, competitors enter, and a SKU you labeled Fix may turn into a Scale. The reserve lets you respond without breaking the plan.
Forecast in Ranges, Then Review Monthly
A single forecast line is false precision. Build three cases instead.
- Base case: last year's contribution margins, current conversion, and modest growth on Scale SKUs.
- Downside: a competitor undercuts your top product, ad costs rise, or a shipment lands late.
- Upside: a Fix lands early, a Scale SKU takes more share, and you have the cash to stock for it.
For each case, record the contribution dollars and the peak cash tied up, not just revenue. Those two numbers tell you whether the plan is survivable and whether it is worth doing.
Then pick the leading signals you will review monthly. We covered these in the metrics that actually predict Amazon growth. When a signal moves, you adjust labels and budgets, not the whole plan.
What to Do This Week
- Export twelve months of SKU-level data and build the contribution margin view, including ads, returns, and storage.
- Write down your cash ceiling and your inventory capacity as two plain numbers.
- Label every SKU Scale, Fix, or Cut, and attach an owner and date to each Fix.
- Draft the budget from those labels, with a reserve you leave unassigned.
- Place your first reorder decisions for the Scale list before supplier lead times shrink your options.
Do this in October and you enter January with a plan built on what your business actually earns, not on a number that sounded good in a meeting.