Most winter footwear orders are placed with a sneaker reflex. The buyer looks at a lined boot, thinks about the sizes that move fastest on an Air Force 1 Low or a Samba, writes 41 to 44 with a bit of 45, and moves on to the next line. The order arrives, the middle sizes go in three weeks, and by mid November the only stock left is 39 and 46, sitting there until the January markdown.
The problem is not the buyer's instinct about which sizes sell. It is that a size curve is a per category, per brand, per model calculation, and winter footwear breaks almost every assumption carried over from sneakers. Different sizing systems, fewer breaks, a sock allowance that shifts the whole distribution, and categories where the customer buys a size block rather than a size. If you are placing winter orders now, this is the part worth getting right before you argue about colourways. For the wider seasonal timing question, we covered it separately in why winter footwear buying starts in summer.
Winter footwear does not use one sizing system
A single winter purchase order can contain four different sizing models. Treating them as one grid is where most size errors start, because the number of variants you have to buy to cover the same range of feet is completely different in each case.
| Sizing model | Typical in the winter mix | Adult breaks to cover the range |
|---|---|---|
| Standard EU run with half sizes | Lined and weatherised sneakers and boots from the big performance brands | 12 to 16 |
| Whole sizes only, sold under both genders | Sheepskin slippers and mules | 8 to 10 |
| Block sizing, one SKU covers three sizes | Apres ski, Moon Boot Icon family | 4 to 5 |
| Single unisex EU scale, width instead of gender | Lined clogs and shearling sandals | 10 to 12, plus a width decision |
Block sizing is the one that changes the maths most. The Moon Boot Icon range is sold in blocks such as 36 to 38, 39 to 41 and 42 to 44, with the brand advising customers to size up if they fall between, so the entire adult market is covered by four or five variants. That means depth per variant is not optional. Each block carries the demand of three sizes, and a single unit per block is a display order, not a sellable assortment.
Whole sizes only works the other way. When a slipper style is offered in whole EU sizes and no halves, every half size customer is pushed to a neighbouring whole size, and the true peak of your curve gets taller and narrower than a sneaker curve for the same customer base. The middle three sizes will carry a much larger share of demand than they would in a run with halves.
Rank sizes by sell-through, not by units sold
The single most common analytical mistake is reading last season's units sold per size as if that were demand. Units sold per size is mostly a reflection of how deep you bought that size. The size you bought six of will always outsell the size you bought two of, and that tells you nothing.
Use sell-through per size instead. Take one model from last winter, in one gender, and lay it out.
| EU size | Bought | Sold | Sell-through | Days to sell out |
|---|---|---|---|---|
| 40 | 2 | 1 | 50% | not sold out |
| 41 | 4 | 3 | 75% | not sold out |
| 42 | 6 | 5 | 83% | not sold out |
| 43 | 6 | 6 | 100% | 31 |
| 44 | 4 | 4 | 100% | 18 |
| 45 | 2 | 2 | 100% | 12 |
| 46 | 2 | 0 | 0% | not sold out |
Read by units sold, 43 and 42 are the stars and 45 barely registers. Read by sell-through and days on shelf, 45 sold out in twelve days and 44 in eighteen, which means both were under-bought and you spent the back half of the season turning customers away in the sizes that actually cleared fastest. The correction is not to move budget into 42, it is to add depth at 44 and 45 and cut 40 and 46 to one unit each or drop them.
Two rules make this reliable. First, only compare sizes within the same model and gender, never across a brand. Second, treat any size that sold out before the halfway point of the selling window as understated demand, and give it more units next time even if its unit count looks small. The full method for reading three weeks of sales before committing budget is in our piece on footwear sell-through rate.
The sock allowance moves the whole curve
Winter customers do not buy the size they wear in summer. Thicker socks, insulated linings and the expectation of wearing a boot for five hours outdoors all push the same foot up. Some brands build that in and their boots run generous, so buyers are advised down half a size. Others run snug when new because the lining has not yet compressed, so buyers are advised up.
The practical effect on your order is that the peak of the curve sits half a size to a full size higher in winter than it does in sneakers, and it sits at a different point for each model depending on how that model fits. This is not something to guess at. Before writing depth, check the brand's own fit guidance for the specific model, then check your own return and exchange data from last winter. Exchanges are the cleanest signal you have. A model where most exchanges went one size up is a model where your curve needs to shift up, and where the size you bought least of is the one customers actually wanted.
Publish that guidance on the product page as well. Half of the exchange volume on winter footwear is avoidable with one line of fit copy, and exchange volume on a boot costs more in shipping than it does on a pair of trainers.
Depth in the core, presence in the tails
A size curve is really two decisions, not one. The core sizes are a depth decision and the tail sizes are a presence decision, and they should be budgeted separately.
- Core, three to four sizes. This is where roughly two thirds of your units belong. Depth here is what lets you survive a peak week without going out of stock in the middle of the run.
- Shoulder, two sizes either side of the core. Buy these to cover, not to trade. One or two units each is usually right on a first order.
- Tail, the extremes. Their job is completeness of the listing, not turnover. One unit, or zero if the brand has a low minimum and you can reorder singles.
The reason to keep the tails on the site at all is that a broken size run reads as a clearance rack. A shopper who lands on a boot listing showing only 39 and 46 available will assume the whole store is picked over, and that impression carries across to your other categories. Keeping one unit in the tails is cheap insurance for the perceived depth of the catalogue, which is the same argument behind brand diversification at the assortment level.
Curve notes by winter category
Each winter category behaves differently enough to deserve its own curve. Working through the main ones:
- Lace-up leather boots. The widest run you will buy and the one closest to a sneaker curve, with the peak nudged up for socks. Half sizes exist and matter here, so do not collapse the run to whole sizes to save on variants.
- Sheepskin slippers and mules. Whole sizes only, sold to both men and women, extremely narrow demand peak. Buy fewer breaks and much more depth per break. This is also the category most likely to sell out entirely in December, so the reorder decision has to be made in November.
- Apres ski and block-sized boots. Four or five variants for the whole adult range, so every variant is a core variant. Order depth on all of them and place the reorder early, since block SKUs are also the first thing the brand runs out of. Stock sitting in the Moon Boot range tends to move in a short, sharp window rather than across the season.
- Lined clogs and shearling sandals. Unisex EU scale with a width consideration instead of a gender split, which means overlapping demand from two customer groups on the same variant. Depth in the overlap zone, roughly EU 39 to 43, is where these sell out. Both lined clogs and the equivalent Birkenstock styles behave this way, and we went through the run structure in more detail for Birkenstock wholesale stock-ups.
- Kids. A separate curve entirely, and not a scaled-down adult one. Kids demand clusters by age group rather than around a mean, and the parent buys one size up deliberately. Never derive a kids curve from your adult data.
What to do before you send the order
- Export last winter's sales at variant level, not product level. If your reporting only gives you product totals, you cannot build a curve and everything above is guesswork.
- Calculate sell-through per size for each model you are repeating, and record days to sell out for anything that hit zero.
- Flag every size that cleared before the season midpoint. Those get more units, regardless of how small their unit count looks.
- Check the sizing model for each new line before writing depth. Count the breaks. If a model has five breaks instead of fourteen, your per variant depth has to roughly triple.
- Pull your exchange data by direction. If exchanges skewed one way, shift that model's curve accordingly.
- Split the order into core, shoulder and tail, and check the split. If the core is not carrying around two thirds of units, the curve is too flat.
- Agree the reorder trigger with your supplier at the time of the first order, not in December. A size curve without a reorder plan is a single bet.
None of this requires new software. It requires variant-level data and the discipline to rank by sell-through instead of by units. The buyers who get this right are not buying better products than everyone else, they are simply buying the right depth of the same products, which is what protects margin when the season compresses into six weeks. Our seasonal overview on how to stock smart for fall and winter covers the assortment side of the same decision, and the timing pressure on Q4 replenishment is in why retailers must prepare for Q4 now.
Related reading
- Winter Footwear Buying Starts in Summer
- Fall Winter Footwear: How to Stock Smart
- Top Footwear Picks to Stock This Autumn Winter Season
- Q4 Dropshipping: Why Retailers Must Prepare Now
- Brand Diversification in Footwear Retail
Winter stock currently available to order is in new arrivals, and the full range is in the main catalogue.
