The term “Demand Sensing” can be used for different things, depending on the context.
This blog aims to clarify any confusion about what we mean by Demand Sensing in SAP Integrated Business Planning for Supply Chain, what value it brings and what the differences are from Statistical Forecasting/Demand Planning.
Why should you be using Demand Sensing? Demand Sensing helps you bridge the gap between planning and execution. Not only is it able to include additional data signals into the weekly forecast, but it disaggregates the weekly forecast to days by recognizing the patterns in your daily demand, resulting in
Does your business concentrate on specific days of the week? Do you want to specify non-working days where no demand is expected? Do you want to get a daily demand forecast to optimize your inventory planning and deployment? Do you want your demand forecast to check and match up to existing open orders? Then Demand Sensing is what you are looking for.
Before we deep dive into the differences between Demand Sensing and Statistical Forecasting in SAP IBP, let’s start by checking what “the internet” means when we talk about Demand Sensing. Here is a very good definition: “Demand sensing is a forecasting method that uses artificial intelligence and real-time data capture to create a forecast of demand based on the current realities of the supply chain. Traditionally, forecasting accuracy was based on time series techniques which create a forecast based on prior sales history and draws on several years of data to provide insights into predictable seasonal patterns. Demand sensing uses a broader range of demand signals, (including current data from the supply chain) and different mathematics to create a forecast that responds to real-world events such as market shifts, weather changes, natural disasters and changes in consumer buying behavior.”
There are multiple points that go into this definition of Demand Sensing:
Is this definition fitting to what Demand Sensing can do in SAP IBP? Yes. Are all of these points prerogative of Demand Sensing in SAP IBP? No.
Artificial Intelligence and demand signals (both internal and external) can be used in SAP IBP both with the multivariate (advanced) forecasting algorithms from Statistical Forecasting and in Demand Sensing.
For the long- and mid-term forecast, the available algorithms are Gradient Boosting of Decision Trees, (Seasonal) ARIMAX and Multiple Linear Regression. All these algorithms are able to include additional demand signals, however they require an adequate of data both in the past and in the future. Typical examples are promotions or holidays, which are known in advance and can help improve the forecast accuracy. Sometimes this type of forecasting is called “Long-term Demand Sensing”. More details on how to use Gradient Boosting can be found in my previous blog.
For the short-term forecast, Demand Sensing algorithms based on Multiple Linear Regression and xGBoost can be used. These algorithms can take into account additional demand signals, which can also be given only in the past horizon.
By definition, current data is only available for the recent past and present, and is only relevant for the short-term horizon. A typical example of this is weather signals: knowing that there is going to be a snowstorm may affect your sales, but you are only going to know this a few days in advance. For such data Demand Sensing is the right thing to use in SAP IBP: it allows you to add signals that are only in the past horizon, can automatically understand the lag between the signal and the effect of the signal on the sales, and apply this learning to the near future.
Statistical Forecasting in SAP IBP can be used to forecast at any time level. However, as most forecasters know, not all forecast levels are equally good in terms of forecast quality. For example, a forecast at monthly level might show seasonality patterns much clearer than a weekly forecast. For this reason, it is usually not a good idea to use Statistical Forecasting for daily forecast: daily demand is usually intermittent and contains a lot of noise. If you need an accurate daily forecast, then Demand Sensing is the best way to go: after a weekly forecast optimization step, the forecast is smartly disaggregated to days by analyzing the patterns of the past few weeks, taking into account workdays, holidays, and open orders.
I will conclude quoting ChatGPT: “Demand sensing is an important tool for businesses looking to improve their forecasting accuracy, reduce costs, and increase customer satisfaction by providing the products they need, when they need them.”
I hope this gave you a better idea of what we mean by Demand Sensing in SAP IBP and how it could benefit you. If you have any questions, do not hesitate to contact me.
If you are curious to know what our customers think about Demand Sensing and integrating additional signals into their forecast, have a look at these:
This blog aims to clarify any confusion about what we mean by Demand Sensing in SAP Integrated Business Planning for Supply Chain, what value it brings and what the differences are from Statistical Forecasting/Demand Planning.
Why use Demand Sensing
Why should you be using Demand Sensing? Demand Sensing helps you bridge the gap between planning and execution. Not only is it able to include additional data signals into the weekly forecast, but it disaggregates the weekly forecast to days by recognizing the patterns in your daily demand, resulting in
- reduced safety stock
- less expeditions.
Does your business concentrate on specific days of the week? Do you want to specify non-working days where no demand is expected? Do you want to get a daily demand forecast to optimize your inventory planning and deployment? Do you want your demand forecast to check and match up to existing open orders? Then Demand Sensing is what you are looking for.
What is Demand Sensing
Before we deep dive into the differences between Demand Sensing and Statistical Forecasting in SAP IBP, let’s start by checking what “the internet” means when we talk about Demand Sensing. Here is a very good definition: “Demand sensing is a forecasting method that uses artificial intelligence and real-time data capture to create a forecast of demand based on the current realities of the supply chain. Traditionally, forecasting accuracy was based on time series techniques which create a forecast based on prior sales history and draws on several years of data to provide insights into predictable seasonal patterns. Demand sensing uses a broader range of demand signals, (including current data from the supply chain) and different mathematics to create a forecast that responds to real-world events such as market shifts, weather changes, natural disasters and changes in consumer buying behavior.”
There are multiple points that go into this definition of Demand Sensing:
- Artificial intelligence
- Demand signals (including data from the supply chain – i.e. internal and external signals)
- Current data (i.e. data related to recent events)
Is this definition fitting to what Demand Sensing can do in SAP IBP? Yes. Are all of these points prerogative of Demand Sensing in SAP IBP? No.
Demand Sensing in SAP IBP
Using Artificial Intelligence and Demand Signals to Forecast – not only in Demand Sensing
Artificial Intelligence and demand signals (both internal and external) can be used in SAP IBP both with the multivariate (advanced) forecasting algorithms from Statistical Forecasting and in Demand Sensing.
For the long- and mid-term forecast, the available algorithms are Gradient Boosting of Decision Trees, (Seasonal) ARIMAX and Multiple Linear Regression. All these algorithms are able to include additional demand signals, however they require an adequate of data both in the past and in the future. Typical examples are promotions or holidays, which are known in advance and can help improve the forecast accuracy. Sometimes this type of forecasting is called “Long-term Demand Sensing”. More details on how to use Gradient Boosting can be found in my previous blog.
For the short-term forecast, Demand Sensing algorithms based on Multiple Linear Regression and xGBoost can be used. These algorithms can take into account additional demand signals, which can also be given only in the past horizon.
Using current data for Forecasts – only in Demand Sensing
By definition, current data is only available for the recent past and present, and is only relevant for the short-term horizon. A typical example of this is weather signals: knowing that there is going to be a snowstorm may affect your sales, but you are only going to know this a few days in advance. For such data Demand Sensing is the right thing to use in SAP IBP: it allows you to add signals that are only in the past horizon, can automatically understand the lag between the signal and the effect of the signal on the sales, and apply this learning to the near future.
Other differences between Statistical Forecasting and Demand Sensing in SAP IBP
Statistical Forecasting in SAP IBP can be used to forecast at any time level. However, as most forecasters know, not all forecast levels are equally good in terms of forecast quality. For example, a forecast at monthly level might show seasonality patterns much clearer than a weekly forecast. For this reason, it is usually not a good idea to use Statistical Forecasting for daily forecast: daily demand is usually intermittent and contains a lot of noise. If you need an accurate daily forecast, then Demand Sensing is the best way to go: after a weekly forecast optimization step, the forecast is smartly disaggregated to days by analyzing the patterns of the past few weeks, taking into account workdays, holidays, and open orders.
Conclusion
I will conclude quoting ChatGPT: “Demand sensing is an important tool for businesses looking to improve their forecasting accuracy, reduce costs, and increase customer satisfaction by providing the products they need, when they need them.”
I hope this gave you a better idea of what we mean by Demand Sensing in SAP IBP and how it could benefit you. If you have any questions, do not hesitate to contact me.
If you are curious to know what our customers think about Demand Sensing and integrating additional signals into their forecast, have a look at these:
- How Does Demand Sensing Accelerate the Digital Journey to a Sustainable Supply Chain? From Arla
- Driving the future of mobility with accurate demand planning – ZF Friedrichshafen
- SAP Managed Tags:
| User | Count |
|---|---|
| 7 | |
| 2 | |
| 2 | |
| 2 | |
| 2 | |
| 1 | |
| 1 | |
| 1 | |
| 1 | |
| 1 |
