Most retailers bet on what shoppers will want a year from now. Zara stopped guessing. It reads what is selling today and restocks twice a week.
This is one of the most quietly powerful operations stories in retail, and in early 2026, it got sharper.
For decades, the fashion industry ran on long forecasts. Designers guessed a year ahead, factories in Asia produced in huge batches, and whatever did not sell ended up on the markdown rack. Big bets, long lead times, high waste.
Zara's parent company, Inditex, built the opposite system and has now wrapped it in what it calls "Quiet AI." No flashy chatbots. Just AI woven into the supply chain itself.
Here is the operations problem it solves:
A) Fashion demand is almost impossible to predict far ahead. Trends move in weeks, not seasons.
B) Before its full RFID rollout, Zara's store inventory accuracy sat around 65%. Roughly one in three items was misplaced, miscounted, or invisible to the system.
C) Forecasting harder does not fix a problem this volatile. Reacting faster does.
How Inditex solved it:
a) Item-level RFID ("soft tagging," built with Intel) now tracks 100% of garments in real time across 5,600 stores.
b) AI demand forecasting drives around 85% of initial production allocation decisions.
c) AI trend detection scans social media, runways, and street style to spot trends 3 to 4 weeks faster than traditional methods.
d) Nearshoring in Spain, Portugal, Morocco, and Turkey keeps design-to-rack time near two weeks.
e) €1.8 billion invested in tech and logistics across 2025 to 2026.
The result is a "test and react" model: make small bets, read real demand, and restock what actually sells.
Three takeaways for operations leaders:
1) When demand is volatile, speed of reaction beats accuracy of forecast. You cannot predict a trend, but you can respond to one.
2) Visibility is the foundation. You cannot react to what you cannot see. RFID fixed the 65% blind spot first, then AI made it smart.
3) Small, frequent bets beat large, infrequent ones. Lower inventory risk, less markdown, faster learning.
If your demand shifted next week, would your supply chain react in weeks, or would it still be working off last year's forecast?
Image credit: RadiusDigital, ShanghaiGarment,
@ETBrandEquity
#SupplyChain #Operations #Zara #Inditex #FastFashion #RFID #AI #DemandForecasting #RetailOperations #Industry40 #InventoryManagement #Nearshoring #OperationsManagement