Poor merchandising and sizing decisions due to unmined exchange/return data

$2-8M annually per 50-100 store district (inventory carrying costs, markdown write-offs, lost sales from stockouts, labor for manual reallocation) • $20-40M annually for a $1B retailer (per search results, 20-50% poor size match vs. 70% best-in-class costs $40M total in lost sales + reduced margins) • $5-15M annually for mid-to-large retailers in special occasion segment (higher margins + higher return rates in this category; 24%+ baseline returns, potentially 30%+ for special occasion)

Customer service forwards return notes in unstructured text; buyers manually count 'too big' vs. 'too small' mentions; occasional ad-hoc calls to RDCs to audit physical returns • District Managers receive aggregate inventory reports but no detail on size-level exchange/return patterns; they manually cross-reference POS sales to RMA data; they make informal requests to buyers ('Hey, we never sell size L in this location, can we get more M?'); buyers ignore or delay due to lack of systematic evidence • Excel pivot tables of aggregate sales data, manual email reviews of customer complaints, isolated notes from customer service, zero systematic mapping of exchange reasons back to SKU/size