USE CASES

Demand Forecasting

Improving forecasting models for a sporting good manufacturer

Poor demand forecasting is a thing of the past. Businesses today are able to harness the power of AI to optimise supply chains, minimise costs, and meet customer expectations effectively. At NEWNOW we have developed demand forecasting models to transform the way in which our clients operate. Here we'll dive into a cockpit we constructed for a sporting goods client.

Context

Previously, the company was relying on manual, Excel-based forecasts with high effort and low accuracy. Planning uncertainty led to both stock-out and excess stock periods. Their large product portfolio with a high number of similar SKUs and frequent changes to product assortment meant it was difficult to track forecast effectively. Strong seasonality (fashion trends) and many external demand factors further complicated the situation.

Collaborating closely with the team, we constructed an ML-model predicting demand by retail store and SKU and factored in the impact of external factors such as search demand and weather.

From here we constructed a cockpit from which our client was able to understand how sales and demand would evolve over time by retail store. Our client was able to quickly respond to demand changes and adjust inventory across stores and experienced a significant reduction in excess stock.

Benefits

Precise demand prediction at every product & retail level
Intelligent product allocation recommendations to reduce stock-out or excess stock risk
Improved resource planning and accelerating decision making from days to minutes