Amazon January restock: why it's a cash decision, not just a stock decision
The order you place in January is made on numbers that have not settled yet, and Amazon's February 1 fee changes raise the cost of getting it wrong. Here is how to size it against your cash.
Every January, the same decision lands on your desk: how much to reorder now that peak is over. It looks like a stock question. How many units, and when.
It is really a cash question. The order you place in January ties up money for months, and this year Amazon is changing what it costs to get that order wrong in either direction. Order too little and you pay one fee. Order too much and, from February 1, you pay more for units that sit too long.
The hard part is that you make the call on numbers that have not settled yet.
Why January numbers mislead you
Three things distort the picture in January, and they all push you toward the wrong order size.
Peak demand is still in your averages. A 30-day or 90-day sales average taken in early January is full of holiday orders. Read it as your normal rate and you order for a December that is not coming back.
Returns are still arriving. Holiday purchases can be returned until January 31, 2027. Some of those units come back sellable and land in your stock count. Others come back unsellable and just sit there. Either way, the on-hand number you see in mid-January is still moving. Our post on the annual holiday return window covers what that does to your books.
Stockout days look like low demand. If a product ran out in December, the days it was out of stock show zero sales. A simple average treats those days as days nobody wanted it, which shrinks the next order and sets up the next stockout.
None of this is a reason to wait until March. It is a reason to know which of your numbers are still moving before you commit cash to them.
Two fees, one on each side of the order
Amazon announced its 2027 US fee changes on November 5, 2026. Most of them start on January 15, 2027, and the storage changes start on February 1, 2027. Two of them bracket your January reorder.
Order too little: the low-inventory-level fee
The threshold stays at 28 days of supply. Since January 2026, Amazon has measured it for each variation on its own, at the seller FNSKU level, rather than across the parent ASIN.
What changes on February 1 is how Amazon checks it. Today the fee looks at trailing 30-day and 90-day sales. From February 1, it applies only when inventory falls below 28 days of supply based on actual shipments in both the four weeks before and the four weeks after the assessment date. Amazon says the aim is to target products that genuinely need replenishing, and to trigger the fee less often on sudden demand spikes and end-of-season or end-of-life products.
The practical point does not change. A product you expect to keep selling still needs more than 28 days of supply on hand.
Order too much: one age-based storage rate
From February 1, Amazon ends the storage utilization surcharge. It also removes the aged inventory surcharge and folds it, with the monthly storage fee, into a single rate based on how long a unit has been stored.
Amazon says inventory that turns in under six months will pay the same or less, and aged inventory will pay more. A January order placed against inflated holiday averages is exactly the kind of stock that turns into aged inventory by summer.
Amazon's announcement does not list the new age-based rates, so compare them with what you pay today, in your own account, before you send a large Q1 shipment.
The cash side nobody plans for
A reorder is not one decision. For most catalogs it is ten or twenty, made one product at a time. Each one can look sensible on its own. Together they can ask for more cash than January has. It is the same trap as running inventory, ads, pricing and margin in separate tools: each piece is right about its own slice and wrong about the business.
And January is a tight month for cash. Holiday refunds keep coming through to the end of the month. Peak fulfillment fees run until January 14. From January 15, Amazon's FBA fulfillment fees rise by an average of $0.19 per unit, and the 3.5% fuel and logistics surcharge stays on top of them. Meanwhile the supplier wants paying for the order you have not sold yet.
So the honest question is not "how much of each product should I reorder?" It is "which reorders can this month's cash actually carry, and which ones can wait?"
That means costing both mistakes. Ordering late costs you sales, rank and possibly the Buy Box, and winning the Buy Box back is not always cheap. Ordering early costs you storage and cash you cannot use for anything else. Neither is free, and the cheaper one is different for every product.
A January reorder routine you can run by hand
You can do all of this in a spreadsheet. It takes an afternoon, and it is worth doing even if you never use a tool for it.
Step 1: Split your products into three groups
Keep selling and reorder. Keep selling but let stock run down. Stop selling. The last group should not get a reorder at all, however good its December looked.
Step 2: Rebuild demand without the holidays
Use last year's January to March sales, or your November average before the Black Friday spike, as the starting rate. Leave out any days a product was out of stock.
Step 3: Wait for the returns to settle where you can
For products with plenty of cover, place the order after the January 31 return deadline, once you can see what came back sellable.
Step 4: Work out days of cover for each product
Units available plus units already on the way to FBA, divided by the daily rate from step 2.
Step 5: Set an order-by date, not just a quantity
Find the day your cover falls to 28 days, then count back your real supplier and shipping lead time. That is your order-by date.
Step 6: Check inventory age before you add more
In your FBA Inventory page, look at how old your current units are. Stock already approaching six months will cost more from February 1, and adding to it makes that worse.
Step 7: Fit the whole list to your cash
Add up every reorder, compare it with the cash you expect in January and February, and decide which orders go now, which shrink, and which wait.
The last step is the one most sellers skip, because it is the one a spreadsheet makes hardest.
Where Inventory Oracle fits
Inventory Oracle is the Sydon AI agent for this decision. For every product, it tells you when to reorder and how much, so you neither run out nor overstock. It refreshes every night, for every product, and puts the result in one plan.
Here is how it maps to the steps.
Honest demand, not a holiday average. Inventory Oracle forecasts demand as a range, labelled with what it is based on: your own sales history, or an average of similar products in your own catalog. It builds in monthly seasonality from your own sales, and it leaves out days a product was out of stock, so a December stockout does not shrink your next order. When there is not enough data, it says so rather than guessing.
Days of cover and an order-by date. For every product you see days of cover, the estimated chance of running out in the next 7 and 30 days, a reorder quantity and the date the order has to go in. It uses your service level and your supplier lead time, learned from your own order history, and subtracts units already on the way to FBA.
Between the two fees. It raises orders that would drop a product below the 28-day low-inventory-fee line, and trims orders that would overshoot into slow-moving stock. It also respects Amazon's capacity limits, and warns you when inventory is aging toward higher storage costs, with markdown, removal and liquidation options priced in dollars.
The right products only. It blocks reorders for listings that are inactive or not buyable. Products that have stopped selling go to an end-of-life review instead of a reorder.
One plan against your cash. This is the step a spreadsheet makes hardest. Inventory Oracle fits every reorder into one plan against your cash projection and costs "too early" against "too late" in dollars. Each reorder is marked funded, partly funded, deferred or dropped, so you see which orders this month's cash can carry. Large orders are paused for you to review before anything else happens.
One thing stays with you on purpose. Inventory Oracle recommends, explains and flags. It never places a supplier order, never creates an FBA shipment and never changes anything on your Amazon account. You place every order yourself, with the plan in front of you.
What this does not do
Four limits, stated up front.
- No orders are placed for you. Inventory Oracle does not place, draft or approve purchase orders, and it does not create FBA shipments. You do that yourself.
- Forecasts are ranges, not promises. They come with their basis shown, and there is no claimed accuracy figure behind them. Read them as a guide, and trust your own knowledge of a product where it says more.
- Amazon US only today. Other marketplaces are not covered.
- It does not calculate fees or margins. Inventory Oracle plans stock and cash, not profit per product.
And a few things Amazon has not said, which no tool can settle for you. The new age-based storage rates were not in the November announcement, so check the current rates in Seller Central before you size a large Q1 shipment. Fee rules can change again, so confirm the details on Amazon's own fee pages before you act on them.
The point of all this is narrow. Not more inventory data, but one reorder plan your cash can actually carry.
Sources
- Chain Store Age: Amazon to increase, streamline shipping and logistics fees in 2027
- PPC Land: Amazon sellers face average $0.19 FBA fee rise from January 15, 2027
- Supply Chain Dive: Amazon to expand fuel and logistics surcharge to more services
- eFulfillment Service: Amazon 2026 FBA fee changes, what sellers need to know
- Amazon Seller Central: 2027 US Referral and Fulfillment by Amazon fees
Build the brand, not the burden behind it.
Amazon, FBA, and Seller Central are trademarks of Amazon.com, Inc. or its affiliates, used here for descriptive purposes only. Sydon AI is not affiliated with or endorsed by Amazon.