Demand Sensing

The core of demand sensing lies in access to details regarding customer behavior and the use of mathematical algorithms that analyze this. Patterns and trends that would otherwise be undetectable become visible. Downstream demand data, such as customer and channel data, is used to identify trends in demand, provide advanced warnings of problems, and reduce the gap between plan and reality in the supply chain. The faster discrepancies can be identified, the faster and smarter the company can react.

Demand sensing imports daily demand data, quickly assesses demand changes compared to an established demand pattern, and evaluates the statistical significance of the change. Through this, demand sensing performs short-term forecast adjustments automatically using probabilistic pattern recognition and predictive analytics.

Examples of demand data that can be used for demand sensing include:

• The next downstream node in your supply chain, for example a distribution center or a warehouse.
• Further downstream nodes, such as POS data from stores, creating abilities to model and control an extended supply chain.
• More advanced sources of unstructured “big data” in demand signal repositories (DSRs), containing, for example, macroeconomic trends, social media, or customer on-line buying patterns.