Mastering Size Curve Analysis: How to Stop Overbuying Fringe Sizes

Why Your Best Seller Still Ends Up on Markdown

Most apparel teams don't have a sales problem. They have a size distribution problem.

A style can post strong sell-through, generate healthy revenue, and still leave inventory sitting in the stockroom months later. The issue is often buried inside the size curve. Core sizes disappear early, fringe sizes hang around, and what looked like a successful buy slowly turns into markdown liability.

Anyone who has worked through an end-of-season review has seen it. The style itself wasn't the problem. Customers wanted it. They just couldn't get it in their size.

This is why size curve analysis deserves more attention than it typically gets. Many retailers still evaluate performance at the style or SKU level. That view can hide some of the most expensive inventory mistakes in the business.

The result is familiar:

• Stockouts in high-demand sizes
• Excess inventory in low-demand sizes
• Lost full-price sales
• Higher markdown exposure
• Inventory dollars trapped in units customers never really wanted

Size curve analysis helps retailers understand demand at a much more useful level. When done properly, it improves inventory productivity, protects margin, and creates better buying decisions season after season.

The Real Purpose of Size Curve Analysis Is Profit Protection

A size curve represents the expected demand distribution across sizes within a product, category, or assortment. Most teams think of it as a merchandising tool. In practice, it is a profitability tool.

Every buy contains an assumption about customer demand. If a retailer purchases a style in a standard size run without validating actual demand patterns, they're assuming every customer behaves the same way.

They don't.

Size demand changes by category, fit, brand positioning, geography, store format, and channel. A relaxed-fit women's knit top may have a very different size profile than a fitted dress. A suburban store may sell a different size mix than a downtown flagship. Ecommerce often behaves differently than physical stores.

The consequences become obvious when inventory is analyzed by size instead of by style.

Imagine a dress that sells through 80% overall. On paper, that sounds healthy. But a deeper look reveals Medium and Large sold out in week four while XS and XXL remained available for months.

The style appears successful in reporting. In reality, revenue was left on the table because customers could not purchase the sizes they wanted.

Why SKU-Level Metrics Hide Size Problems

Most planning reviews focus on metrics such as:

• Total sell-through
• Revenue
• Gross margin
• Weeks of supply (WOS)
• Inventory turn

Those metrics are valuable, but they rarely expose size-level imbalance.

A planner reviewing only style performance may conclude the buy worked. Meanwhile, store associates have been hearing the same complaint for weeks:

"Do you have this in a Medium?"

The answer is often no.

That missing visibility creates a feedback problem. Future buys get built using sales history that already reflects stockouts. The same size mistakes are repeated because planners are learning from distorted data rather than actual demand.

Over time, the organization becomes very good at forecasting what it managed to sell, not what customers actually wanted to buy.

The Hidden Cost of Overbuying Fringe Sizes

Fringe sizes are not the enemy.

Retailers need assortment breadth. Customers expect size inclusivity. The challenge is determining how much inventory belongs in each size.

Problems emerge when fringe sizes consume inventory investment that should have been allocated elsewhere.

Most excess inventory situations begin long before product reaches stores. They start during the buy.

A planner may use historical pack ratios, vendor recommendations, or legacy size curves that haven't been challenged in years. The assortment looks balanced. The math works.

Months later, the consequences show up:

• Core sizes sell out early
• Replenishment opportunities are missed
• Remaining inventory becomes concentrated in fringe sizes
• Markdown pressure increases
• Inventory productivity declines

The frustrating part is that stockouts and excess inventory often originate from the same decision.

Why Stockouts and Markdowns Come From the Same Decision

Retailers frequently treat stockouts and overstock as separate problems.

They're often two sides of the same size allocation mistake.

Every extra unit bought in a slow-moving fringe size consumes inventory dollars that could have supported a proven demand size.

Think about a women's denim program where size 28 and 29 consistently drive demand. If too much inventory is committed to less productive sizes, there is simply less capacity available for the sizes customers are actively searching for.

The outcome creates a double penalty.

First comes the lost revenue from stockouts in core sizes.

Then comes the markdown cost associated with excess fringe inventory.

Neither outcome is desirable. Together, they create a margin leak that can persist for an entire season.

This is why size curve planning should be viewed as margin management, not just assortment planning.

Building Demand-Based Size Curves Instead of Assumption-Based Curves

Many retailers are still operating with size curves created years ago.

The business has changed. Customer behavior has changed. The curves often haven't.

Effective size curve analysis starts with actual demand, not historical assumptions.

That distinction matters because sales data can be misleading.

If a Medium sells out after three weeks, recorded sales only reflect inventory availability. The sales report doesn't show all the customers who would have purchased that size if inventory had remained available.

This is one of the most common planning traps in apparel.

A stockout gets mistaken for weak demand because the data collection process ends the moment inventory disappears.

Clean the Data Before Building Curves

Not all historical sales data should influence future size curves.

The strongest size curves are typically built from:

• Full-price selling periods
• In-stock weeks
• Comparable products
• Recent customer demand patterns
• Similar fit and silhouette characteristics

Planners should also separate analysis by meaningful demand drivers.

Examples include:

• Category
• Gender
• Fit profile
• Silhouette
• Sales channel

A slim-fit men's woven shirt may require a different size distribution than a relaxed-fit woven shirt, even within the same category.

Applying a single curve across both products creates avoidable imbalance.

This doesn't mean retailers need hundreds of unique curves. That quickly becomes unmanageable.

The goal is identifying differences that materially affect inventory outcomes.

A practical approach is usually better than a perfect one.

A retailer with ten well-maintained demand-based curves will often outperform a retailer with one generic curve applied across the entire assortment.

A Common Example From Apparel Planning

Consider a retailer launching a new oversized sweatshirt collection.

Historical data from fitted sweatshirts suggests a traditional size curve. The planning team uses it without adjustment.

Within a few weeks, Large and XL sizes are selling significantly faster than expected. Smaller sizes remain available.

Nothing unusual happened. Customers simply preferred the oversized fit and intentionally sized up.

The product wasn't forecast incorrectly.

The size demand was.

Without size-level analysis, that lesson is easy to miss.

Why One National Size Curve Usually Fails

Uniform size curves make operations easier.

Unfortunately, customers don't organize themselves around operational convenience.

One of the biggest planning mistakes is assuming every store should receive the same size distribution.

Different markets often exhibit different demand patterns.

Age demographics, income levels, local preferences, tourism, climate, and channel mix can all influence size demand.

A store serving young urban professionals may require a different size profile than a family-oriented suburban location.

Yet many retailers continue allocating inventory using one national curve.

The result is predictable.

Some stores run out of key sizes while others accumulate excess inventory in the exact same style.

From National Curves to Store Clusters

Most retailers don't need store-specific curves for every location.

The operational complexity becomes difficult to manage.

A more practical solution is cluster-based planning.

Examples include:

• Urban vs suburban stores
• Regional clusters
• Climate-based clusters
• Store volume clusters
• Ecommerce vs store inventory pools

This approach improves demand alignment without creating excessive planning overhead.

Allocation becomes more accurate because inventory arrives closer to expected demand patterns.

Transfers decrease.

Reactive firefighting decreases.

Store teams spend less time searching for inventory that should have been allocated correctly in the first place.

A retailer may discover that stores in one region consistently sell larger size profiles while another region skews smaller. Once that pattern is visible, allocation decisions become much easier.

In many cases, better size allocation creates a larger financial impact than incremental forecast improvements.

Turning Size Curve Analysis Into an Ongoing Planning Process

One mistake I see repeatedly is treating size curves as a preseason exercise.

The buy gets finalized, inventory ships, and the size curve disappears until next season's planning cycle.

That's too late.

The best retailers use size curves as a living planning input throughout the season.

They influence:

• Initial buys
• Open-to-buy decisions
• Allocation strategies
• Replenishment priorities
• Inventory transfers
• End-of-season reviews

Demand changes. Trends emerge. Products behave differently than expected.

The goal is not to create a perfect curve before launch.

The goal is to continuously improve decisions as new information becomes available.

Monitor the Right Size-Level Metrics

Most inventory dashboards still focus heavily on style-level performance.

That view should be supplemented with size-level metrics, including:

• Sell-through by size
• WOS by size
• Stockout rates by size
• Markdown rates by size
• Size realization rates

These metrics often reveal issues weeks before traditional reporting surfaces them.

A growing imbalance in week three can still be corrected.

A growing imbalance discovered after markdowns begin is usually much harder to fix.

This is where technology can help. Many planning teams still rely on spreadsheets to monitor size-level performance across thousands of SKUs. The process becomes difficult to maintain as assortments grow.

Platforms that continuously monitor demand signals and inventory positions can surface size-level risks much earlier than manual reviews, allowing planners to act before stockouts or overstock become expensive.

The objective isn't automation for its own sake. It's giving planners enough visibility to make better decisions while there is still time to influence the outcome.

The Retailers That Improve Fastest

The retailers with the strongest inventory productivity are rarely the ones with perfect forecasts.

Perfect forecasts don't exist.

The advantage usually comes from identifying demand signals faster and responding sooner.

A planner who notices a size imbalance after three weeks still has options:

• Adjust replenishment
• Reallocate inventory
• Modify future buys
• Prioritize transfers

A planner who notices after end-of-season markdowns has already absorbed most of the financial impact.

Size curve analysis provides an early-warning system for those decisions.

That is where much of its value comes from.

Conclusion

Overbuying fringe sizes is rarely caused by forecasting failure alone.

More often, it stems from outdated assumptions about size demand.

Retailers that consistently achieve stronger full-price sell-through tend to approach sizing differently. They analyze demand at the size level, account for stockout distortion, localize allocation decisions, and monitor size-specific performance throughout the season.

The payoff is tangible.

Fewer stockouts in core sizes.

Less inventory trapped in fringe sizes.

Lower markdown exposure.

Better inventory turns.

Stronger margin retention.

At its core, size curve analysis is not about selling more units. It's about placing inventory where demand actually exists and avoiding the costly habit of investing inventory dollars where demand doesn't. In apparel retail, that distinction often separates productive inventory from expensive inventory.