Introducing Demand Sensing

Objective

After completing this lesson, you will be able to use SAP IBP's demand sensing for optimized daily forecasts.

Introduction to Demand Sensing

In this process, optimized daily forecasts are created for multiple products in the background in a highly automated way for a short period between 4-8 weeks. The forecasts are based on a review of the consensus demand and the most recent demand signals retrieved from internal sources such as sales orders and external sources such as social media or competitors. Information regarding promotion and new product introduction can also be taken into account.

The process results in adjusted short-term forecasts that can be used as optimized input by processes such as operational supply planning or transportation planning.

Demand Sensing helps you to sense your daily demand for a short term horizon such as 4-8 weeks. It is usually run weekly or daily (sometimes even sub-daily) to always incorporate the latest demand drivers such as new sales orders. Based on a more accurate, and a larger number of input factors, demand sensing can reach a much better forecast accuracy in the short term horizon, than the traditional demand plan. The more accurate sensed demand plan is, the better it can then be used to react to shifted priorities that are predicted for the near future already a little ahead of time to ensure constant, good customer service levels, minimize stock-outs and to avoid costly fire drills.

Key Learning Points and Benefits for a Supply Chain Planner

Improved short-term forecast accuracy, improved forecast bias, improved safety stock calculations, improved visibility for required deployment changes, which leads to reduced stock out scenarios.
Use cases for demand sensing are wide spread, e.g.: company struggles with volatile markets and demand shifts, company receives orders from their customer and want to include them into the short-term planning horizon. Company has already applied demand driven business strategies, but wants to further improve. Company is capable of adjusting their production or transportation plan within a short-term horizon and wants to further improve on this aspect. The company wants to close the gap between a monthly demand planning cycle and daily or weekly demand requirements for the short-term horizon.
Before Demand Sensing weekly forecast of 40 units and equal distribution to both distribution centers East DC and West DC.
After demand sensing the company still forecasts a weekly demand of 40 units but distributes it to the distribution centers matching the sales trends in East and West DC.

Based on the different planning horizons, the following cases need to be considered:

  • Demand sensing creates a short-term demand plan. The reaction to this plan is not part of the SAP IBP solution but is highly dependent on your industry, production boundaries, business strategy, and overall possibility to react in the short-term, as follows:

    • Adjust deployment or transportation decisions.

    • Decide on stock transfers (redistribution) from one Distribution Center (DC) to another for the finished goods in case you see short term regional shifts in demand.

    • Change your production sequence (if possible) in case you see high demand for specific products.

  • In the wholesale/distribution business, material purchasing could also play a role when it comes to demand sensing, as follows:

    • Rework your purchases from external and internal suppliers to procure material easily.
Demand Sensing does not replace the typical demand planning process or the statistical forecasting, as companies still need mid- to long-term demand plan for tactical and strategic purposes. Instead, demand sensing uses the mid- to long-term Demand Plan as a key input factor and improves it in the short-term horizon. Demand Sensing adjusts the mid- to long-term forecast based on certain patterns, brings additional demand signals into consideration, and breaks it down to the daily level.
Demand Sensing — Properties and Algorithms

Compared to classical statistical forecasting, Demand Sensing has the following properties:

  • The demand sensing algorithm is primarily used for demand forecasts for a shorter period such as four to eight weeks, based on the consensus demand and the most recent demand signals retrieved from the ERP system, for example, the quantity in open sales orders. Typically, demand sensing is executed for multiple products in the background and results in automatic adjustment of the short-term demand.

  • Two types of Demand Sensing algorithms exist:

    1. Demand Sensing (Full)
    2. Demand Sensing (Update)

  • Demand sensing is typically run daily. It is executed at product-location-customer level, and the sensed demand is created at the daily time granularity for the time horizon defined in the forecast model.

  • Since the optimization processes can take some time and the optimized weights do not change much from one day to the next within a week, it is recommended to set the system to run the demand sensing (full) algorithm weekly and the demand sensing (update) on a daily basis.

Log in to track your progress & complete quizzes

Learning

SAP FacebookSAP YoutubeSAP LinkedIn