Extending Your Size Curve

Adding new sizes—like moving from a S-L range to XS-XL—is an exciting way to capture untapped demand for a popular product. However, placing that first Purchase Order (PO) can feel like guesswork.

Use this checklist to build a data-backed size curve for your extended range.

1. Do you have a “Lookalike” product?

The most reliable predictor is a similar item already sold in the extended range. For example, if your Navy Sweater already sells in XS-XL, use its curve for your new Burgundy Sweater PO.

Validate the match by comparing common sizes: If the sales distribution for S, M, and L is nearly identical between the two products, you can safely assume the demand for XS and XL will follow suit.

📌 Pro Tip: If a single product comparison feels too narrow, use the aggregate size curve of an entire category (e.g., “All Men’s Sweaters”) to smooth out anomalies.

2. Apply Relative Ratios

If you don’t have a direct comparable, use “rules of thumb” to tweak your existing data. Rather than adopting a generic “standard curve” from the internet, apply ratios to your actual sales.

Example: The “Half-Size” Rule Suppose a standard benchmark says XS usually sells 50% as much as Small.

• Current Sales: S: 100 units | M: 400 units | L: 500 units
• Calculated XS: 50 units (50% of your actual Small sales)

This method preserves the specific “shape” of your existing customer base (e.g., if you lean heavily toward Large) while providing a realistic estimate for the new sizes.

3. Account for Cannibalization

When you introduce edge sizes, they rarely represent 100% “new” demand. Instead, they often “steal” sales from adjacent sizes:

• XS will take some customers who previously settled for S.
• XL will take some customers who previously squeezed into L.

The Strategy: Err on the side of caution for your first order. Shave a few percentage points off the “fringe” sizes and redistribute them toward the middle.

Final Curve Comparison:

• Calculated Curve: XS:4% - S:8% - M:32% - L:40% - XL:16%
• Conservative Curve: XS:3% - S:7% - M:37% - L:39% - XL:14%

The Bottom Line

Predicting size curves is part art, part science. When in doubt:

1. Use comparables whenever possible.
2. Use ratios to keep the curve grounded in your specific brand’s data.
3. Be conservative on the edges to avoid overordering until you have more hard data.