Why a Holiday Forecast Should Start From Net Demand, and What That Changes in the Buy
Consider a cashmere sweater that sells at full price on December 20. The shopper returns it on January 14. The store takes approximately nine days to process it, and the sweater goes back on the floor in the middle of clearance, where it sells at 30% off. While it sat in the returns bin, the replenishment system saw an empty size on the shelf and ordered another unit.
December sales reports count the sweater as a success, while the margin reports that follow show a sale that was provisional, a return which arrived after the season ended, and a retailer that paid to supply one customer’s demand twice.
This is a commercial risk that deserves more attention in discussions about holiday returns. A sale can look final at the register while the demand behind it is still open, and the units that come back arrive on a different clock from the one that planned the buy.
This article treats holiday returns as a planning question. It sets out three measurements that turn a gross demand forecast into a net one, and it shows what changes in buy depth, replenishment and clearance when a retailer plans on that basis.
A Holiday Sale Stays Provisional Until the Return Window Closes
The register records gross sales, while the business keeps net demand, which is gross demand less the units that come back. Retailers expect that gap to be large in the holiday season. In its most recent survey of retailers, NRF and Happy Returns found they expected shoppers to return 17% of holiday sales, and NRF now says that the vast majority of returns arrive in January, after the holidays.
The 17% comes from a retailer survey, and actual rates differ by category, channel and price tier. It still describes the shape of the problem. A large share of holiday demand stays provisional into late January, and a forecast built on register sales treats it as settled in December.
Returns Arrive on a Different Clock From Sales
A unit sold on December 20 and returned within a month lands back in the building in January, after the weeks when it was most likely to sell again at full price. Return policies stretch that clock further. Walmart's 2026 window covers purchases from October 1 and runs to January 31, 2027, so an October sale can come back four months later.
Planning systems follow the sale clock. A unit in transit back from a customer, or waiting for inspection, shows as an empty space on the shelf, so the system sees a gap the retailer has already filled. The size of that mismatch varies by retailer, and by category. And each retailer's own sales, returns and inventory records will show it.
Three Measurements Turn Gross Demand Into Net Demand
A retailer can build the bridge from gross to net with three measurements, each taken by category and channel.
- Return rate: The share of units sold that come back, by category, channel and price tier. A single company average overstates some lanes and understates others.
- Time to return and time to resale: The days from sale to return, then from return to a sellable unit on the shelf. Together they decide how much of the selling window a returned unit still has.
- Recovery price: The price at which the returned unit sells next, compared with its original price and after processing cost.
The first measurement shapes buy depth, the second shapes where stock sits and how much capacity the January floor needs, and the third shapes when markdowns begin.
The Same Unit Can Earn Very Different Margins on Its Second Sale
Take an item priced at $100 with a $55 cost. It earns $45 of gross profit at the register. If the buyer returns it, the retailer refunds $100, spends $8 on processing and sells the unit again, and the outcome depends on when and at what price that second sale happens.
The register shows $45 on December 20. After the return, the same unit ends anywhere from $37 to a $23 loss. The figures use assumed costs and describe no retailer's actual returns economics, and some returned units cannot be resold at all. They show how much the timing of the second sale can matter.
Returned Inventory Is Also Supply
The sweater in the opening scenario was already in the building when the system ordered another one. Units in transit or in the returns queue are supply, and a replenishment plan that leaves them out buys stock the retailer already owns. Routing matters as much as counting. A returned unit can go to the store with the strongest demand, to the online channel or into a hold for the next season, and the default of restocking at the original location may not be the best choice.
The same logic applies to the initial buy. If a large share of a category’s holiday units will return, a buy based on gross demand runs deeper than the business needs. Because the returned units can serve part of the January and February demand. A shallower first buy, paired with forecast of when units return, can lower the stock exposed to clearance. This holds only where returned units come back sellable and in season, and each category has to show that with its own history.
Test the Opportunity Before Changing the Buy
Retailers can start with this season's data. Between December 26 and the end of February, they can record the three measurements for each category, so the 2027 plan begins with observed return rates, return timing and recovery prices by category.
From there, a retailer can pick a few categories with stable return histories, size next year's buy on net demand, and compare net gross profit and ending inventory against similar categories that keep the gross-demand buy. Gross sales would hide the effect the test is meant to find.
Return policy belongs in the same conversation. A longer window is a customer experience decision, and it also moves the date when returned units re-enter the business. Planners who know that date can build it into the forecast.
The Question to Ask Before the Next Buy
A holiday sale is worth what the retailer keeps after the return window closes. Gross sales show how much the retailer sold in December, and net demand shows how much it needs to buy, hold and mark down. Before a retailer sizes the next holiday buy on gross sales, the question to ask is how much of this demand it will keep, and when the rest will come back.
The right holiday buy starts with knowing how much demand will stay. Discover how Impact Analytics help retailers connect demand, inventory and pricing decisions at www.impactanalytics.ai.





