Choosing which sizes to offer is a product decision. Choosing how many units to buy in each size is an inventory decision. Clothing brands often combine the two and assume an equal size split, which can create stockouts in core sizes and dead inventory at the edges.
Define the fit before the size label
Small, medium, and large are not universal measurements. Start with the garment's intended silhouette and actual measurements. A relaxed unisex shirt, fitted women's top, oversized hoodie, and compression legging require different grading logic and customer expectations.
Create a measurement chart
Document points of measure for every size: chest, body length, sleeve, waist, hip, inseam, rise, or whatever is relevant. The grade between sizes should be deliberate. This should be incorporated into the tech pack and validated through samples.
Do not assume the same size curve for every product
A size distribution that works for one silhouette may not work for another. Customer gender mix, geography, fit preference, category, and marketing audience can all shift demand. A heavily oversized product may create different buying behavior than a fitted product.
Use actual order data as soon as you have it
Early orders are valuable. Track the percentage of units sold by size for each style, not just overall. Also track how quickly sizes sell out. If medium repeatedly sells out while extra-small remains in stock, future orders should reflect that information.
Separate sales mix from constrained sales
If a size was out of stock for half the launch, its sales share understates true demand. Look at when each size was available. Waitlists, restock notifications, customer-service messages, and returns can provide additional clues.
Offer a broader size range carefully
Inclusive sizing can expand the addressable customer base, but every additional size creates more SKUs and inventory complexity. That is not an argument against broader sizing; it is a reason to plan it intentionally, validate fit, and allocate enough inventory for the sizes offered.
Use smaller test runs when data is weak
If you do not know the correct size curve, a smaller initial order can be more valuable than maximizing unit discount. The goal is to learn without trapping too much cash. Review MOQ strategy when factory minimums make testing difficult.
Returns are sizing data
Track whether customers return because an item runs large, small, short, long, tight, or loose. A high return rate may be a fit-design issue rather than a size-run allocation issue. Product pages and size guides should set expectations clearly.
Plan reorders around lead time
The longer it takes to replenish inventory, the more buffer you need in high-demand sizes. A short lead time lets you operate leaner and correct forecasts faster.
Size planning becomes more accurate with every launch. The goal is to replace assumptions with real demand data while keeping enough flexibility to improve fit and serve more customers over time.