What are Size Curves and How to Accurately Calculate Them

Size curves, sometimes known as size breaks, are a big or a small problem, depending on your ability to forecast demand. Certain products have multiple variable attributes, which means that forecasting for them in a “one size fits all” manner can mean you completely miss the mark. This in turn, causes overstocks, stockouts and eventually, missed sales and disappearing customers.

So, what is a size curve, how do you calculate them, how are they influenced and what role does inventory planning software play in making size curves a small problem or even an asset for your business?

Table of Contents

What is a Size Curve?

A size curve guides you to the ideal ratio of sizes, based on a particular color choice or style. Basic demand forecasting using WOS (weeks of supply) and other less accurate formulas can leave you with the right quantity of overall inventory but the wrong quantity for individual sizes.

Retailers might decide on a 1:2:2:1 (S,M,L,XL) ratio for a particular set of t-shirts without properly analyzing historical data or accurately forecasting demand, which causes stockouts and overstocks. Accurate size curve calculations can fix that.

Get the All-in-One Inventory Planning Software

Forecast demand, issue and track POs, re-order on autopilot, and step up your reporting game across multiple channels and locations. Get in touch to see how Singuli can help you optimize your inventory.

Types of Size Curves

There are two main size curve types: normal distribution and skewed (empirical).

Normal Distribution Size Curve

A normal distribution size curve doesn’t look at historical data. The middle sizes are at the center and bought in higher quantities than the small and large sizes, which are on the outside of the curve.

Skewed (Empirical) Size Curve

Skewed (or empirical) size curves take account of your sales velocity, which is the total number of sales divided by the number of days in stock, for each SKU (stock keeping unit). This gives you a better idea of which sizes are selling so you can base your size curve on more accurate information.

They can often look completely different to the standard “bell shaped” size curves that you see assuming a normal distribution. Although this gives you better accuracy, you need to make sure you’re taking every influence on your size curve into account, which is easier to do with robust inventory planning software.

📌Get Started: With Singuli’s Advanced Forecasts, it’s easy to analyze historical data, market trends and customer inputs, which saves you or your inventory planner time and increases forecast accuracy.

How to Calculate Size Curves

There are three main approaches to calculating size curves:

Raw Sales Ratio

The easiest way to calculate a size curve is using raw sales ratio. You look at the number of units sold within a specified period of time for a particular color choice. Let’s say you sell a total of 350 pairs of pants and 75 of them are in size XL. To calculate the ratio, divide the total number of XL pants by the total pairs of pants and multiply it by 100.

$$\frac{\text{75 pants in size XL}}{\text{350 pairs of pants}} = 21.45\%$$

That means 21.43% of your total sales came from size XL and therefore, 21.43% of your next pants intake should be in size XL too.

The main problem with using a raw sales ratio is that you’re not taking into account all of the data. What effect does selling products at a marked down price have, did stockouts cause the data to be skewed, do you really have enough information to make a sound judgment?

Collection or Category Ratio

One way to improve the ratio is to base your analysis on a larger collection. Instead of taking a sample of 350 pairs of pants, you can look at all slim fit pants instead. This gives you a much larger collection size to work with. Now your data is based on 1,200 slim fit pants and 200 of them are in size XL. As you can see, this changes the percentage of XL pants to 16.67%.

$$\frac{\text{200 slim fit pants in size XL}}{\text{1,200 pairs of slim fit pants}} = 16.67\%$$

You’ve probably already noticed that this still isn’t as accurate as it could be and it still isn’t considering all the data, so what do you do?

Stockout and Promotion Adjusted Ratio

For the optimal size curve, you need to look at the ratio on the weeks where every size is known to be in stock and where there are no promotions for any of your sizes. You can do this with both smaller and larger collection samples. A mathematically sound way of doing this takes a Bayesian approach, which allows you to combine data from different sources in a principled way.

Technically, all of these curves are called populations and samples. The more data you have to work with, the better your result will be.

💡Pro Tip: Automatically compute the right size curve for every sized-based product in your assortment with Singuli’s size-based forecasting and planning tools.

Size Curve Influences

You’ve already seen that there are different types of size curves but what factors influence the end result?

Customer Demographics

Customer demographics have the biggest influence on size curves. Certain populations will have larger, taller, thinner or smaller populations, in general. This can affect the type of products customers are interested in, as well as the sizes they need for the best fit, which will of course, affect your demand planning.

📌Get Started: Every product has its own story. However your products are growing, whenever there’s a change, Singuli’s forecasting models will identify it and adapt accordingly.


If nobody else is selling XXL, XXXL etc., but there is a demographic of people who need those sizes, you’ll probably sell more of those SKUs. Some retailers only focus on the better sellers and don’t carry much stock in the less-sold sizes but that doesn’t mean it isn’t the right SKU to sell for your business.


Suppliers have minimum order quantities to manage, which means they can be restrictive on the sizes they sell. You might need a ratio of 1:2:1:2:3 (S,M,L,XL,XXL) but your supplier needs to sell more size S and size L, so you have to buy a ratio of 2:2:2:2:3 to get the best prices. It stops them from having to sell certain sizes at a discounted rate and shifts that responsibility to you.

Calculate Size Curves Automatically

With all of the information you’ve just read, you might be wondering how to calculate size curves and to anticipate demand for every SKU, even new products. The easiest way is to let Singuli do it for you. Singuli takes your data and automatically turns it into fully formed predictions that you can use for inventory forecasting.

Book a demo to find out more about how it can work for your business.

Size Curves: Get the Best Fit

You know what a size curve is, the different types that exist, what influences them and how to calculate them so now it’s time to use Singuli to find the best fit.

Get the All-in-One Inventory Planning Software

Forecast demand, issue and track POs, reorder on autopilot, and step up your reporting game across multiple channels and locations. Get in touch to see how Singuli can help you optimize your inventory.

Your Message is Sent

Thanks for reaching out. We'll be in touch soon!

Contact Us

We'd love to show you how inventory optimization could transform your business.

Who is Singuli
We’re a multidisciplinary team of engineers, ph.d. researchers and data scientists with decades of retail experience.
Benjamin Kelly, Ph.D
CEO & Co-Founder
Thierry Bertin-Mahieux, Ph.D
CTO & Co-Founder