← Back to Insights
Retail

Demand forecasting that lifted margin by 12%

Case study · BIGFAT AI LABS

A retailer replaced spreadsheet forecasts with a demand model wired into purchasing — cutting stockouts and markdowns and lifting gross margin by 12% across two seasons.

The problem

Buying decisions ran on last year's numbers and gut feel. The result was the classic retail squeeze: lost sales from stockouts on the winners, and margin-killing markdowns on the losers.

What we built

A forecasting model that blends sales history, seasonality, promotions and local signals, feeding store- and SKU-level predictions directly into the replenishment workflow buyers already use.

The outcome

Stockouts on top sellers dropped, end-of-season markdowns shrank, and gross margin rose 12% across two seasons. Buyers spent less time arguing over numbers and more time acting on them.

Why it stuck

Forecasts are monitored against actuals every week and the model retrains as new data lands. Accuracy is visible to the team, so confidence grew with each cycle instead of eroding.

Put these ideas to work.

Talk to us