Skip to content
Contact us
  • There are no suggestions because the search field is empty.

Size Curve Calculations

In retail, understanding customer preferences is crucial for optimizing inventory management. One key concept that helps retailers align their stock levels with actual customer demand is size curves.

Size curves provide insights into how demand is distributed across different sizes for a specific product category in each store. By leveraging size curves, our customers ensure the right products are available in the right sizes, at the right locations.

What Are Size Curves?

Size curves represent the relative demand for each size of a product, showing how much weight or importance each size holds in relation to overall sales at a given store. These curves are based on estimated demand, meaning they reflect the actual purchasing behavior of customers for each size by estimating how many units could have been sold even when no stock was available, rather than relying on historical sales levels alone.

By default, size curves are calculated at a store-family-size_set level, which means the following factors are considered:

  • The store location
  • The product family
  • The specific size sets within that family

This level of detail ensures that retailers can tailor their inventory strategies to the unique needs of each store and product type.

How Are Size Curves Used?

Once size curves are established, they are applied to the overall product-store forecast in order to break it down and obtain a forecast at SKU-Store level. This process provides a more accurate prediction of demand for each individual size, helping to optimize stock levels and prevent unnecessary overstocking or even stockouts.

Ensuring Accuracy and Granularity

To ensure robust size curves, a minimum volume of sales is required for each store-product category-size set combination. This threshold ensures that the size curve calculations are based on sufficient data to accurately reflect customer preferences.

As a consequence, if there is not enough sales data at the most granular level, the system will use broader store categories to calculate the size curves. This flexibility allows the model to generate reliable forecasts, even in cases where sales data for individual sizes may be limited.


Size Curves Fallback System Overview




  • Level 0 (Highest Granularity)
    • STORE DIMENSION: Individual Store
      • The size curve is calculated for a specific store.
    • PRODUCT DIMENSION: Family
    • SIZE SET DIMENSION: Individual Size Set
      • The system analyzes the demand for each size in the product category within that store.
  • Level 1
    • STORE DIMENSION: Region / Area
      • If there isn't enough data at the individual store level, the system uses aggregated sales data within the store category chosen by the customer.
    • PRODUCT DIMENSION: Family
    • SIZE SET DIMENSION: Individual Size Set
      • The system analyzes the demand for each size within the product category across the region.
  • Level 2
    • STORE DIMENSION: Store Network
      • When regional data is not sufficient, the system expands to the entire store network, analyzing sales across all stores.
    • PRODUCT DIMENSION: Family
    • SIZE SET DIMENSION: Individual Size Set
      • The system analyzes the demand for each size within the product category across the network.
  • Level 3 (Lowest Granularity)
    • STORE DIMENSION: Store Network Without Robustness Criteria
      • As a final fallback, the system looks at the entire store network without applying minimum robustness criteria. This means the data may not meet the usual thresholds for reliability, but it ensures that size curves can still be calculated.
    • PRODUCT DIMENSION: Family
    • SIZE SET DIMENSION: Individual Size Set
      • The system analyzes the demand for each size within the product category across the network, even if the data is less reliable.

Why Nextail’s Size Curve Methodology Outperforms Traditional Replenishment Models

Many retailers are accustomed to replenishing based on the immediate past sales of specific sizes. For example, if a store sells two units of size XL in a given week, they might expect two more units of XL to be sent from the warehouse. While this method appears straightforward, it can lead to suboptimal stock levels and long-term inefficiencies, especially when it comes to sizes with sporadic demand.

At first glance, reacting directly to the sales of individual sizes may seem like an efficient way to meet customer demand. However, this approach doesn't consider the broader trends or the true customer buying patterns for each size at each store. This is where the size curve methodology offers a more sophisticated, data-driven solution that leads to improved inventory management and better sales outcomes.


Here’s why our size curve methodology is far superior:

1. Sporadic demand vs long-term patterns
  • A traditional approach might replenish two units of XL simply because two were sold last week, without considering whether XL is typically a low-demand size for that store. This often leads to overstocking of sizes that have only experienced temporary demand.
  • Our size curve methodology, however, looks at long-term demand patterns. If XL is historically a slow mover in that product category and store, the system prioritizes replenishing sizes that consistently perform well, ensuring that your inventory is better aligned with ongoing demand.

2. Preventing Casual Sales from Skewing Inventory Decisions
  • In many stores, there are core sizes that represent the bulk of the sales. Traditional systems can easily overlook these if short-term sales of other sizes spike unexpectedly. This may lead to stockouts of your best-performing sizes.
  • The size curve model prevents these anomalies from distorting your stock levels. Instead, it focuses on the sizes that truly drive performance at each location. By doing so, you minimize the risk of ending up with excess inventory that is less likely to sell, tying up valuable shelf space.

Conclusion

While it might seem intuitive to replenish sizes based solely on the previous week’s sales, this approach can easily result in overstocking or stockouts, both of which harm store profitability. Our size curve methodology, on the other hand, uses a more strategic, long-term approach, relying on accurate demand insights to deliver smarter replenishment decisions.

By focusing on the true buying behavior at the store-family-size set level, we help you ensure that each store has the right mix of sizes based on long-term trends—not just short-term fluctuations. This leads to higher sales, lower inventory risks, and a much more efficient use of your stock.