What happens when one of retail's most stable planning inputs quietly stops behaving the way it always has?
Size curves don't usually make headlines. They sit in the background of planning decisions, quietly shaping how retailers buy, allocate, and replenish inventory season after season. Most planning teams don't revisit them often because, historically, they didn't need to. That's starting to look like a mistake.
A recent Impact Analytics report, tracking apparel demand shifts shows size distribution moving toward smaller bands across multiple categories. Larger sizes are not collapsing but losing share slowly and consistently, in ways that most planning systems are not designed to catch early enough to act on. The report puts more than 400 million apparel units at risk of misalignment, with over $5 billion in inventory value and margin exposure if retailers keep planning against historical size assumptions.
The Size Curve Is Beginning to Behave Differently
Size curves rarely break all at once. What’s happening now is more of a gradual drift, and drift is harder to manage than a sudden break because it doesn’t set off any alarms. Categories still perform well enough, and inventory still moves, so the signal hides inside what looks like normal seasonal variation until the gap between what was planned and what sold becomes too large to ignore.
A single percentage point decline in the large size share across the U.S. adult apparel market shifts roughly 120 million units annually.
When the Small Shifts Start Moving to the Middle
When smaller sizes gain share and larger sizes soften, the curve doesn't just rebalance at the edges. The pressure eventually reaches Medium, the size that carries the heaviest volume weight in most assortments.
A 0.75 percentage point decline in the Medium share alone translates to roughly 90 million units shifting bands annually.
Current data shows Medium holding nationally but softening in markets where GLP-1 adoption is running ahead of the broader curve. That geographic unevenness is what makes it particularly hard to plan around. A national average can look stable while specific markets are already experiencing meaningful compression.
If the signals were already there, why weren’t planning systems reading them that way?
Planning systems were built around the assumption that body-level demand changes slowly, giving retailers time to observe and adjust. GLP-1 adoption and related structural changes in consumer body composition are compressing that timeline considerably.
The Curve Was Already Incomplete Before GLP-1
Here is a dimension of the problem that rarely gets discussed. The historical size curves retailers rely on were built from what sold, not from what customers actually wanted. When a size runs out early in the selling window, that unmet demand disappears from the data entirely. Just a clean sell-through that gets read as success and fed back into the next buy as validation. The curve gets built on a foundation that was missing some of its most important information.
Stockouts in high-demand sizes don’t just represent lost sales. They silently distort the size curve and represent missed revenue that never appears in any report.
If smaller sizes have been stocking out faster than planning teams realized, the true scale of demand shift has been understated for longer than the GLP-1 conversation suggests, with 1.1 million metric tons of CO2 and 98.5 million gallons of water annually accumulating as the cost of that misalignment.
Building a more accurate curve requires reconstructing demand from the ground up at store level, something no buying team can do manually across hundreds of locations and thousands of SKUs.
The Curve Looks Different Depending on What You Are Selling
Updating a size curve is a start, but it only solves the problem if size is the only dimension shifting, and the Impact Analytics latest research shows it isn’t. Color, fit, silhouette, and finish are each behaving differently across size bands. Slim and athletic silhouettes are concentrating volume heavily in XS and S, while even relaxed fits, which distribute more evenly, are still gradually migrating in the same direction.
When size and attribute are modeled independently, misallocation can exceed five percentage points within a single subclass.
The Channel Gap Planning Systems Are Missing
The degree of divergence between store and digital has widened considerably, and the pace is accelerating.
Current data shows physical stores migrating toward smaller size bands materially faster than digital channels. That gap matters because most planning systems still aggregate curve data across both, which means the store-level shift gets diluted when viewed at a combined level. The result is a curve that looks more stable than it is in the places where inventory is physically sitting.
National averages are now actively masking local overexposure in stores.
A buyer working from a total company selling data has no visibility into which stores ran out of which sizes or how early those stockouts occurred. Reconstructing that signal requires a bottom-up approach across every location, well outside what any buying team can manage. Every season that gap goes unmodeled is another buy cycle where store-level inventory gets committed against a curve that aggregate performance has artificially inflated.
The Reset Starts This Planning Cycle
The planning decisions being made right now are already running alongside a shift that has been building for several seasons. Every forward buy commitment built on 2023 size distributions adds to an imbalance that is visible in stores, across channels, and across attributes simultaneously. The Impact Analytics 2026 analysis puts 130 to 210 basis points of gross margin growth and a 3 to 5% reduction in markdown sales on the table for retailers who get size precision right.
That requires modeling channel curves independently, integrating size and attribute planning, and elevating size precision to the same leadership dashboard as markdown rate, inventory turn, and return ratio.
The $5 billion question is whether planning systems will reset before the next buy cycle locks in yesterday’s assumptions.
Access the full Impact Analytics 2026 analysis to explore the complete data behind the size curve shift, what it means for assortment planning, and where the margin opportunity sits for retailers who move first.
