Seasonal demand planning on Amazon means estimating how much each product will sell in each part of the year, then working backwards to set stock, ad budgets and inbound deadlines. You do it in four steps: pull your own sales history, adjust it for events and growth, convert the result into stock and budget by week, and build in contingency time for things going wrong. If you are reading this on 22 October 2026, the first hard deadline for Black Friday stock is only days away, so the checklist further down comes first.
TL;DR
- Forecast by week, not by quarter. Q4 behaves differently in October, November and December, and each major event needs its own goal.
- Treat inbound dates as hard deadlines. This year, stock for Black Friday readiness needs to arrive at UK fulfilment centres by 28 October, and a festive peak fee applies from 15 October 2026 to 14 January 2027.
- Start your own data work two to three months before the peak, and keep the model no more complex than the data you can actually record.
- Plan ads and stock together. Budgets should rise with demand, and ads should stop when stock runs out.
- Add contingency weeks to every lead time. Customs delays and carrier problems are what turn a good forecast into a stockout.
Table of Contents
- Why seasonal demand needs its own plan
- The dates that matter for Q4 2026
- Step 1: Build your baseline from your own history
- Step 2: Map goals event by event
- Step 3: Turn the forecast into inbound stock
- Step 4: Plan advertising around the same forecast
- Our perspective: forecasts fail on margin, not just on units
- Automating the routine parts
- A short checklist for this week
- FAQ
- Sources
Why seasonal demand needs its own plan
An average month tells you very little about a seasonal product. A product that sells 100 units a week for ten months and 600 a week for two is not a 200-a-week product, and ordering as if it were leaves you empty at the worst moment and overstocked in January.
The spikes are real and sharp. In the UK, shoppers spent £1.3 billion online over the two days of Prime Day 2024, up 8.8% year on year, according to Adobe’s analysis. A spike of that kind rewards sellers who had stock in place and ad budget ready, and it punishes those who noticed it on the day.
There are two ways to get seasonal planning wrong:
- Understocking. You run out, lose Buy Box and ranking momentum, and pay for ads that have nothing to sell.
- Overstocking. You tie up cash and pay storage on units that will not sell until next year. Amazon’s own guidance on supply chain and inventory makes the same point: optimisation is about avoiding both stockouts and overstock.
Good planning sits between the two, and it is mostly arithmetic you can do yourself.
The dates that matter for Q4 2026
Start with the calendar, because it fixes your deadlines. Amazon’s UK peak season 2026 post in Seller Forums sets out the key points:
| Item | Detail |
|---|---|
| Black Friday readiness, inbound deadline | Inventory must arrive at UK fulfilment centres by 28 October to be Prime-badge eligible |
| Festive peak fulfilment fee | Applies from 15 October 2026 to 14 January 2027 |
| Fee level, small and standard parcels | Around £0.12 per unit on average |
| Fee level, large and extra-large envelopes | Around £0.07 per envelope |
Two details are easy to miss. The deadline is for stock to arrive, not to ship. A pallet that leaves your supplier on 27 October is not ready. And the peak fee is small per unit but it is not small on a thousand units, so it belongs in your margin calculation before you set a promotional price.
Deadlines move from year to year, so check the Seller Central page for the current dates before you commit to a shipping plan.
Step 1: Build your baseline from your own history
The most established forecasting method is also the simplest: look at what you sold last year and the year before, week by week.
- Export weekly units sold per ASIN for at least the last 12 months, ideally 24.
- Strip out distortions. Remove weeks when you were out of stock, ran an unusual discount or had a one-off bulk order. Those weeks understate or overstate real demand.
- Find the seasonal shape. Divide each week’s units by the average weekly units for the year. A week at 2.5 means demand ran at two and a half times normal.
- Apply the shape to a new baseline. If your current run rate is 80 units a week and the week before Black Friday has historically been 2.5 times normal, plan for around 200.
- Adjust for growth. If you have added reviews, ad spend or variations since last year, your baseline has moved. Use recent weeks, not last year’s absolute numbers.
The seasonal index in step 3 is the core of the whole exercise. Once you have it for one product, you can compare it with similar products to sanity-check your numbers.
A worked example
Suppose a pet accessory sells 60 units a week in September, and last year the pattern was:
- October: 1.2 times September
- Black Friday week: 3.0 times September
- Early December: 2.0 times September
- January: 0.7 times September
If September 2026 is running at 70 units a week, your plan is roughly 84 in October, 210 in Black Friday week, 140 in early December and 49 in January. That is a working forecast, not a promise. The next steps turn it into something you can order against.
Keep the model honest
It is tempting to reach for sophisticated methods. Research on inventory forecasting does show a shift toward machine learning: one systematic review found that 43.6% of recent studies use neural networks or deep learning, and that simple methods such as moving averages stay useful as baselines to beat. Logistics practitioners cited by Amazon add a practical warning: do not design a model more complex than your data recording can support. For most sellers, a clean weekly history and a seasonal index will beat a clever model fed with patchy data.
When you do compare approaches, use plain error measures. MAE, RMSE and MAPE are the standard ones, and MAPE (average percentage error) is the easiest to read. Test your forecast on last year: apply your method to 2024 and see how close it gets to 2025.
Step 2: Map goals event by event
Amazon Ads guidance recommends that seasonal planning starts at least two to three months before the peak and that you set specific goals for each major event, not one for Q4 as a block. Shopping behaviour changes a lot between October, November and December.
A simple way to do it is a table with one row per event window:
| Window | Typical aim | What to plan |
|---|---|---|
| Prime Big Deal Days (October) | Early gift buyers, ranking momentum | Deal eligibility, first stock wave |
| Pre-Black Friday build | Reviews, traffic, audience warm-up | Ad budget ramp, listing checks |
| Black Friday week | Volume | Peak stock, bid caps, margin floor |
| Cyber Monday to mid-December | Gift shoppers, delivery-date sensitivity | Replenishment, last order dates |
| Late December to January | Clearance, returns | Reduced ad spend, avoid overstock |
For each window, write down the units you expect, the stock you need on hand at the start, and the advertising budget you are prepared to spend. Decide your target profit in advance. Event pricing and extra fees can turn a high-volume week into a loss-making one if you only watch revenue.
Step 3: Turn the forecast into inbound stock
A forecast only matters if it becomes stock in the right place at the right time. Work backwards from the date you need units available for sale.
Use real lead times and contingency weeks
Amazon’s own freight team says planning for the busiest months starts well before the season, and recommends building contingency weeks into your supply chain calendar rather than assuming perfect transit. Sellers who forecast on past sales alone, without supplier capacity and lead times, are the ones who hit stockouts in Q4. The contingency should cover customs delays and carrier problems, not only normal transit.
A practical lead-time formula:
Order date = date stock must be available for sale − Amazon check-in time − transit time − production time − contingency
If you have never measured your check-in time, measure it this year. Check-in is where many sellers are caught out in November and December.
Respect the check-in squeeze
Fulfilment centres prioritise customer orders in November and December, so inbound receiving times can stretch. Amazon’s seller forum post suggests that sellers who wait until the last minute risk delays that leave inventory unavailable during peak traffic. One tactic it mentions is booking night and weekend delivery appointments to avoid weekday congestion.
Know your capacity before you order
There is no point forecasting 3,000 units if your FBA limits allow 1,800. Check capacity early. Our guide to calculating FBA headroom and avoiding blocked shipments shows how to work out what you can send. Amazon has also noted in the seller forums that stock moved from Amazon Warehousing & Distribution into FBA is not subject to the standard FBA capacity limits, which some sellers use as a workaround near their ceiling. Confirm how it applies to your account before relying on it.
Oversized or heavy items also cost more than many sellers expect. Check how the 5,000 dimensional weight rule affects UK FBA fees before you commit to a packaging size for a big order.
If you also sell through your own store and share stock across channels, a separate inventory management optimisation setup can help you keep Amazon and direct stock from competing for the same units.
Step 4: Plan advertising around the same forecast
Stock and ads should rise and fall together. Amazon Ads’ seasonal guide, drawing on experienced sellers, says the common error for new holiday sellers is underestimating lead times. It also suggests launching holiday campaigns soon after Halloween, running an always-on approach and reviewing bids on Fridays and Mondays to catch shifting behaviour during the rush. You can read it in Amazon’s three-step guide to seasonal sales.
The same source notes that Amazon Advertising offers budget rules, which raise campaign spending automatically during peak periods so campaigns do not run out of budget early.
Three habits protect your margin:
- Set budgets by window, not by month. Use the event table above and allocate spend to Black Friday week separately. Our guide to portfolio budgets on Amazon Ads covers how to group campaigns and cap spending in the UK.
- Pause ads on products that are about to run out. Spending on a listing you cannot ship wastes money and damages performance. Here is how to pause Amazon ads when a product is out of stock.
- Check for lost Buy Box. Ads on a listing without the Buy Box convert poorly. See how the Buy Box affects your Amazon ads and spend.
After the peak, audit what the spend actually achieved. Attribution lags, so the numbers you see on day two are incomplete. Use our weekly ad audit approach, with a 7 to 14 day attribution wait, and check for ad cannibalisation using 30 days of TACoS so you know whether ads created sales or just replaced organic ones.
Our perspective: forecasts fail on margin, not just on units
Most seasonal advice stops at units: how many to order, when to ship. In our experience working with Amazon sellers’ data, the more damaging mistake is forecasting revenue and discovering in January that profit was far thinner than expected. Peak fees, higher advertising costs, discounts and returns all arrive at once, and a spreadsheet that tracks sales alone hides them.
So we suggest forecasting three numbers per product and window: units, ad spend and profit after fees. If the third number is negative at your planned price, no amount of stock will fix it. We built Osellpa around that idea. It connects to Amazon’s API so profit, fees and advertising sit in one place, and automation handles the repetitive parts of PPC so your bids and budgets keep up with changing demand while you deal with suppliers. Sellers using it have reported sales increases of up to 20%, though results depend on the account, the products and the season.
Another point we would stress: data that you can trust beats a forecast that looks clever. Record stockout days, promotion dates and price changes every week. In a year you will have the clean history that makes the method in this guide work.
— Harry
Automating the routine parts
Once your plan exists, the work is monitoring it. Sellers who manage this by hand usually end up with several spreadsheets that disagree. Osellpa brings profit tracking, automatic and manual PPC optimisation, inventory management and alerts into one tool, with pricing based on your Amazon sales volume:
| Plan | Monthly | Annual |
|---|---|---|
| Launch | £10 per month | £96 per year |
| Scale | £20 per month | £192 per year |
| Advanced | £40 per month | £384 per year |
You can compare the plans on the Osellpa pricing page. If your main concern is stock, our article on automating replenishment without letting FBA fees eat your margins shows what to automate and what to keep manual.
A short checklist for this week
It is 22 October, so if you are late, do these in order:
- Check what is already in transit and confirm it will arrive by the 28 October deadline if you want Black Friday readiness.
- Add the festive peak fee of around £0.12 per small or standard unit to your margin calculation for the period to 14 January 2027.
- Set your minimum acceptable profit per product for Black Friday week and stop ads or promotions that fall below it.
- Review ad budgets for the next six weeks by window.
- Schedule a Friday and Monday review of bids and stock cover until the new year.
- Note everything: stockout days, price changes, promotions. Next year’s forecast starts here.
For next year, begin the full cycle in late summer, and take a close look at your product mix with a product research workflow built for Amazon UK so you are not building a seasonal plan around a product with weak fundamentals.
FAQ
When should I start seasonal demand planning for Amazon Q4?
Amazon Ads guidance says to start at least two to three months before the peak. For a Black Friday in late November, that means beginning your data work in August or early September, with supplier orders placed soon after. Starting later is possible but leaves little room for delays.
What are the peak months for Amazon in the UK?
The heaviest trading falls in November and December, around Black Friday, Cyber Monday and Christmas, with Prime Big Deal Days in October opening the season. Summer Prime Day is the other major spike, and the 2024 event alone saw £1.3 billion in UK online spend over two days. Your own sales history will show whether your category peaks in the same weeks.
Do I need AI to forecast seasonal demand?
Not to start. A weekly sales history and a seasonal index give a usable forecast, and simple averages remain a standard baseline in research. AI methods can improve accuracy when you have enough clean data, but a complex model on poor records will mislead you.
How do I avoid overstock after the peak?
Plan the January run rate as carefully as the peak, since Amazon notes that overstock brings unnecessary storage fees. Set a stock cover target per product, hold back reorders that rely on optimistic forecasts and use the post-peak period to clear slow lines.
Sources
- Peak season 2026: inventory dates, peak fees, deal cutoffs and promotions (Amazon Seller Forums)
- Amazon Seller Forums: Peak season alert and FBA cutoff
- Adobe: Record breaking Prime Day provides welcome boost to UK retailers
- LACCEI paper on inventory forecasting methods
- Amazon Ads: 3 steps to help boost seasonal sales
- Amazon Ads: Global shopping events guide
- Amazon Freight: Peak season inbound freight tips
- Amazon UK: Managing supply chain and inventory logistics
