Inventory forecasting and safety stock for seasonal peaks
Ready for a 3PL? Compare 200+ fulfilment providers instantly, or let our experts find your perfect match.
Share
Inventory forecasting is the process of predicting how much of each product you'll sell over a given period, so you can order the right amount at the right time. Safety stock is the extra stock you hold on top of that forecast to protect against demand spikes and late deliveries. Get both right before a seasonal peak and you sell through your bestsellers without ending January buried in stock.
Peaks make forecasting harder. Demand rises sharply, promotions distort normal patterns, supplier and freight lead times stretch, and your 3PL's goods-in team is busy with everyone else's stock too. A forecast built on an average month will miss all of that.
This guide covers how inventory forecasting works, how to forecast for a seasonal peak step by step, how to calculate safety stock with worked examples, how to adjust it for peaks, and how to plan peak stock with your Third-Party Logistics (3PL) provider.
What is inventory forecasting?
Inventory forecasting uses past sales, planned activity and market signals to estimate future demand for each Stock Keeping Unit (SKU). The forecast feeds your purchase orders, inbound deliveries and the stock levels you plan with your 3PL.
These are the main methods, and you can combine them:
Method | How it works | Best for |
Historical trend | Uses recent sales, adjusted for growth | Steady sellers with plenty of history |
Seasonal index | Compares each period's sales with the average to find seasonal patterns, then applies them to the current trend | Products with a predictable peak, such as gifting or summer lines |
Promotional uplift | Adds the expected effect of discounts, campaigns and events | Black Friday, sales and launches |
Like-for-like | Uses a similar product's history as a stand-in | New products with no sales history |
Bottom-up | Builds the forecast from channel or customer-level plans | Brands selling across several channels or wholesale accounts |
Forecast at SKU level where you can, then check the total makes sense against your overall revenue plan. A style-level forecast that looks right can still hide the wrong size or colour mix.
How to forecast inventory for a seasonal peak
Start planning peak stock early. For products made overseas, that can mean placing orders several months before the peak arrives.
1. Clean last year's data
Pull weekly sales by SKU for the same period last year. Remove distortions such as weeks when a product was out of stock, one-off bulk orders and promotions you won't repeat. Stock-outs are the big one: if a bestseller sold out halfway through peak, last year's sales understate real demand.
2. Apply this year's trend
Adjust last year's peak by how the business has grown since. If sales over the last three months are running 20% ahead of the same months last year, that's a reasonable starting point for peak growth, before any changes to your promotions.
3. Add your promotional plan
List every planned promotion, email campaign, creator collaboration and marketplace event, with dates. Estimate the uplift for each from similar past campaigns. A deeper discount or a bigger audience than last year means a bigger uplift.
4. Plan new products separately
For products with no history, use a similar product's sales as a guide and keep the forecast conservative, with a plan to reorder or move stock quickly if it sells faster than expected.
5. Build the forecast by week
Peak demand isn't spread evenly. Black Friday week, Cyber Monday and the last shipping days before Christmas can each carry far more volume than an average week. A weekly forecast tells you when stock needs to be at your 3PL, as well as how much.
6. Work backwards to order dates
For each SKU, take the week’s stock that needs to be available to sell, then subtract your 3PL's receiving time, freight time and supplier lead time. That gives you the latest order date. Add a buffer, because lead times tend to stretch in the months before peak.
7. Review and update weekly
Once peak starts, compare actual sales with the forecast every week, or every day during Black Friday week. Reorder or reallocate stock early for lines running ahead, and slow down marketing on lines that are running low.
What is safety stock?
Safety stock is a buffer of extra units held above your expected demand. It covers two kinds of uncertainty:
Demand uncertainty: customers buy more than forecast, for example because a product goes viral or a competitor sells out.
Supply uncertainty: stock arrives later than planned because a supplier, freight route or goods-in team is running behind.
Safety stock works alongside the reorder point, which is the stock level that tells you it's time to order again:
Reorder point = (average daily demand × lead time in days) + safety stock
The right amount of safety stock is a trade-off. Too little and you run out of bestsellers at the worst moment. Too much and you tie up cash, pay for extra storage and risk markdowns after peak. The goal is to hold enough to hit the service level you've chosen for each product, and no more.
How to calculate safety stock
There are several ways to calculate safety stock. The three below cover most ecommerce situations, from a quick estimate to a method that allows for unreliable lead times.
All three use the same example product: a bestselling jumper that sells an average of 40 units a day, with a supplier lead time of 25 days.
Method 1: the max-average method
Safety stock = (maximum daily sales × maximum lead time) − (average daily sales × average lead time)
If the jumper's best day was 70 units and its longest lead time was 35 days:
(70 × 35) − (40 × 25) = 2,450 − 1,000 = 1,450 units
This method is quick and needs no statistics, but it assumes your worst sales day and your longest delivery happen together. It usually produces a large buffer, so it suits low-volume products or brands just starting to plan safety stock.
Method 2: the standard deviation method
Safety stock = Z × standard deviation of daily demand × √lead time
Z is a number linked to your target service level: the chance of not running out during a lead time.
Service level | Z |
90% | 1.28 |
95% | 1.65 |
98% | 2.05 |
99% | 2.33 |
If the jumper's daily sales have a standard deviation of 12 units and you want a 95% service level:
1.65 × 12 × √25 = 1.65 × 12 × 5 = 99 units
The reorder point is then (40 × 25) + 99 = 1,099 units.
This method gives a much smaller, more targeted buffer. It's a widely used approach for steady sellers with a reliable supplier, and spreadsheets can calculate the standard deviation for you.
Method 3: allowing for variable lead times
Safety stock = Z × √((lead time × standard deviation of daily demand²) + (average daily demand² × standard deviation of lead time²))
If the jumper's lead time varies by a standard deviation of 5 days:
1.65 × √((25 × 12²) + (40² × 5²)) = 1.65 × √(3,600 + 40,000) = 1.65 × 208.8 = about 345 units
The result is more than three times Method 2, and almost all of the increase comes from lead time uncertainty. That's typical in the run-up to peak, when freight and supplier delays are more likely. Use this method when lead times are unpredictable, which is often the case for products shipped from overseas.
Adjusting safety stock for seasonal peaks
A safety stock figure calculated on an average month will be too low for peak. Adjust it in four ways.
Use peak-period demand: calculate average daily demand and its standard deviation from last year's peak weeks, adjusted for growth, instead of a full-year average. Demand is higher and more volatile during peak, so both numbers rise.
Use peak lead times: suppliers, freight and goods-in teams are all busier before peak. Ask your suppliers and 3PL for realistic peak lead times, and use the longer figure.
Set service levels by product: you don't need 99% availability on everything. Rank SKUs by margin contribution and set higher service levels for your bestsellers and hero products, with lower levels for the long tail. Our guide to SKU rationalisation covers how to rank your range.
Plan when to stop reordering: once peak passes, you'll want safety stock to run down instead of being replenished. Set a date for each seasonal SKU after which you stop reordering, so you don't end January with a full buffer of products that have stopped selling.
Recalculate safety stock before each peak, and again once early sales show whether the forecast is running high or low.
Working with your 3PL on peak stock
Your forecast is only useful if your 3PL can receive, store and ship the stock it calls for. Share it early, and plan these points together.
Share the forecast: give your 3PL weekly volume forecasts by SKU, promotional dates and expected order peaks. It uses those numbers to plan staff, space and carrier capacity, so update them if your plans change.
Agree inbound cut-offs: ask for the last date stock can arrive and still be ready to sell before each peak event, and how long receiving is taking. Book deliveries early, because goods-in is under pressure in November.
Plan the space: safety stock takes up room. Check your 3PL has space for your peak stock levels, and how extra storage is charged. Our platform data shows pallet storage pricing varies more between UK 3PLs than most other fulfilment costs, so the cost of holding a larger buffer depends heavily on your provider.
Consider overflow or a second site: our platform data shows many UK 3PLs run more than one warehouse. A second site can hold overflow stock or serve a different region, and some brands use a second 3PL during peak. Our guide to managing a multi-3PL network covers how to make that work.
Get accurate stock data: you need real-time stock levels from your 3PL's Warehouse Management System (WMS) to track sales against the forecast and trigger reorders on time.
Plan for returns: returns after peak add stock back into the warehouse. Fast processing gets it back on sale during the January sales instead of after them. Our guide to ecommerce returns management covers the process.
Our Black Friday checklist covers the wider set of questions to agree with your 3PL before peak.
Common forecasting mistakes
Watch out for these when you plan peak stock.
Forecasting from stocked-out sales: if a product sold out last year, its sales figures only show what you had in stock, which understates real demand. Estimate the lost sales and add them back.
Using one safety stock rule for every product: a flat percentage buffer overstocks slow sellers and understocks bestsellers. Set levels by product, based on demand variability and importance.
Ignoring lead time risk: lead times are often the biggest source of uncertainty before peak, as the worked example above shows. Track actual lead times by supplier and use them in your calculations.
Forgetting goods-in time: stock that's arrived but hasn't been received and checked can't be sold. Build your 3PL's receiving time into every order date.
Planning by month: a monthly forecast hides the weeks that matter most. Plan peak stock week by week.
Not updating the forecast: a forecast set in August and never revisited misses everything that happens next. Review it weekly once peak approaches.
No plan for leftovers: decide before peak how you'll clear excess seasonal stock, whether through January sales, outlet channels or bundles, so you're not making those decisions under pressure.
Inventory forecasting and safety stock FAQs
What is inventory forecasting?
Inventory forecasting is predicting how much of each product you'll sell over a period, using past sales, seasonal patterns, planned promotions and growth, so you can order the right stock at the right time.
How do you calculate safety stock?
A widely used formula is: safety stock = Z × standard deviation of daily demand × √lead time, where Z reflects your target service level, such as 1.65 for 95%. A simpler alternative is (maximum daily sales × maximum lead time) − (average daily sales × average lead time).
What is a reorder point?
The stock level at which you place a new order. It's calculated as average daily demand multiplied by lead time in days, plus safety stock.
How much safety stock should I hold for peak?
It depends on how variable demand is, how reliable your lead times are and the service level you want for each product. Recalculate using peak-period demand and peak lead times, and set higher service levels for bestsellers than for slower lines.
How far ahead should I forecast for Black Friday and Christmas?
Far enough ahead to cover your longest supplier and freight lead times, plus your 3PL's receiving time. For overseas-made products, that can mean finalising orders several months before peak.
What service level should I aim for?
Aim higher for bestsellers and hero products, and lower for slow-moving lines. Each step up in service level needs disproportionately more safety stock, so be selective.
Find a 3PL that's ready for your peak
Peak stock planning comes down to a weekly forecast built on clean data, safety stock calculated for peak demand and peak lead times, and a 3PL with the space and processes to receive and ship it on time. Compare 3PLs on fulfilment.com and get quotes from providers with the capacity your peak needs.
