Business planning has the task of making forecasts of the company’s future based on assumptions that are as realistic as possible. Predictive methods from statistics are a valuable tool here.
An orientation to the future is in the nature of economic planning. It is necessary to identify the goals of the company as a whole or of the individual company divisions and to define how these goals are to be met operationally.
In the traditional descriptive or visual techniques of analysis, results are processed manually in creating forecasts. With the SAP Analytic Cloud Predictive Planning and Forecasting environment, however, progressive model-based approaches are used.
By using mathematical forecasting models, previously undiscovered knowledge and connections can be uncovered both in existing data (samples) and in new data (forecasts). This data can then be used to make forecasts about future events.
There are many advantages in using predictive planning and forecasting. With a better understanding of past developments, future developments can be anticipated more quickly and more reliably. This of course presupposes that the relevant historical data is available and the forecast models are closely integrated in the planning tool. The information acquired is then used to automate the forecasts. This saves significant time, which can be used for other tasks. The use of SAP Analytic Cloud Predictive Planning and Forecasting thus enables a more intensive engagement with value-creating areas and complex interconnections: tasks which require experience, intuition and creativity, skills that machines do not possess. By identifying concrete measures, the future can be actively shaped, and to the company’s advantage.
The predictive functions of the SAP Analytics Cloud (SAC) are designed for use by specialist departments. The individual functions can be used without any detailed knowledge of the underlying statistical processes or algorithms being used. This means that users are guided through the predictive workflows and don’t need to have any predictive skills themselves. Results are presented clearly so that they can be easily interpreted and included in SAP Analytics Cloud Stories and plans.
A new predictive scenario can be created using the Smart Predict environment of the SAP Analytics Cloud. In the main menu select the path “☰ – Create – Create Predictive scenario”.
The SAP Analytics Cloud offers three selection options for creating a new predictive scenario (see figure):
- Time series
The classification process allows objects to be assigned to a particular class based on their attributes.
For example: Which of my customers will respond positively to my marketing campaign?
2. Time series
Time series are used to predict the development of a variable in the future.
For example: What quantity of products should be produced in advance?
Regression is used to find relationships between variables that describe events. For each new event, you obtain an estimate of the value of target variables.
For example, you can predict the price of a house by looking at similar houses.
SAP Analytics Cloud offers an integrated environment that you can use to create predictive models and train them based on existing data. The predictive models thus created can then be used for new data sets and integrated into planning processes.
The use of predictive models enables further automation of planning processes. The result is greater efficiency than with the purely manual approach in traditional planning. In addition, when you use mathematical forecast models you can also increase the quality of the planning values you obtain. The SAP Analytics Cloud gives the planner access to established tools. This eliminates the need for time-consuming integration of external tools, making a more data-powered value calculation possible.
It’s important to remember, however, that SAC forecasting tools do not eliminate the need for the human planner.
Combine the advantages of human and machine to achieve better results. Humans can be used in those areas where creativity and interpretation are necessary. Machines can then be used where clear rules exist for forecasts. Use the information you acquire to evaluate other possible areas of use and expand your know-how.