A sale that comes back isn't demand — it's a round trip. If your forecast counts the sale but ignores the return, every returned unit quietly pads your next purchase order. In categories like apparel, where a third of units can come back, that's a serious over-buy.
Say you sold 100 units of a jacket and 30 came back. Real demand was 70. Forecast on the gross 100 and you'll order as if 100 people wanted to keep it — then do it again next season, compounding the excess. Returns don't just cost you the refund; they distort every future buying decision.
The fix is returns netting: for each product, subtract returned units from the sales that generated them, so the forecast learns from kept units only. A partial return of 4 of 10 units becomes demand of 6. A full return removes the line entirely.
A return usually happens weeks after the sale. You can net it in the return's period, or against the original sale — both are defensible, and either beats ignoring it. What you must avoid is letting a heavy-return month read as negative demand and confusing the forecast; a good implementation floors demand at zero for any period once returns are netted.
Fopsis nets returns automatically for Shopify, WooCommerce, Amazon, QuickBooks, Lightspeed and JD Edwards — handling partial and full refunds, and even credit/return lines from ERP systems. Returned units are subtracted from demand, and monthly demand is floored at zero, so returns never inflate your buy plan.
Returns overstate demand; stockouts understate it. Correcting only one leaves your forecast biased. See why stockouts wreck your forecast for the other half.
Fopsis nets returns out of demand across every channel, automatically.
Start Free Trial