Category-level demand forecasting improves supply chain planning

An automated, data-driven forecasting solution generates real-time category level insights at weekly frequencies to optimize inventory

Business Scenario

The client is a leading American consumer health company. The demand planning team was responsible for setting the organization’s sales, margin and inventory forecasts. Their planners were working in isolation using manual spreadsheet based approaches that were time-consuming, prone to error and subjectivity. The rudimentary methods resulted in inaccurate estimates across the cold, cough and pain categories. The planning team had very little control over the forecast parameters, time horizon for forecasting, cost and explainability.

Sigmoid Solution

We created a customized demand forecasting solution using advanced ML algorithms to estimate future demand for the OTC product categories. Automated data engineering pipelines were operationalized to seamlessly manage all stages of data processing, modeling data preparation, forecast model design, and model selection. Large volumes of data from multiple internal and external data sources such as IQVIA offtake data, 3rd party tool data for pricing, competitive activity, marketing data, POS data, out of stocks etc. were ingested and processed for desired level of forecast granularity– weekly for each category.

Download case study

Business Impact

The automated category demand forecasting solution streamlined the demand planning process, created a unified approach for all planners and significantly improved forecasting accuracy.