What is Predictive Analysis?
In simple terms, the use of historical data to arrive at a possible future event is predictive analysis. It uses statistics & modeling techniques to determine relationships & patterns in historical data to arrive at future indicators. It is used as a decision-making tool.
“Most of the world will make decisions by either guessing or using their gut. They will be either lucky or wrong.” – Suhail Doshi, CEO, Mixpanel.
Organizations can avoid guessing by using predictive analysis, which is backed by data.
SAP Analytics Cloud offers business intelligence, planning, and predictive capabilities in a single cloud solution. In this blog, we will discuss the types of predictive analysis possible in SAC.
Why does a business need Predictive Analysis?
Every business in today’s world needs to have a proactive approach towards consumer demand rather than a reactive one. Let us consider a car manufacturer. First, they want to fulfill all customer orders on time, but they also do not want to keep huge inventories as they will not make a profit till a car is sold. If the vehicle resides in the warehouse, it cannot be counted as part of revenue. So, how does the manufacturer know the number of cars to be manufactured? They predict it! An organization’s current actions are triggered by what they expect or want in the future, a future that they predict.
Following are some cases where predictive planning is used:
- To know the future demand of a product in a manufacturing firm.
- To manage the staffing needs for a services firm to provide an optimal customer experience.
- To plan inventory levels for a retail firm.
- To predict customer churn for a cellular service provider.
- For expense and cost planning scenario in any organization.
Predictive Analytics in SAC
SAC provides predictive analytics capabilities for business users using its own inbuilt machine learning algorithm catering to one of the following methods:
1. Predictive Planning in time series/ line charts
Here, any indicator you show in a time series or line chart can be forecasted. You can use either the Automatic Forecast option or select from one of the Linear Regression or Triple Exponential Smoothing forecast options.
2. Predictive Planning in tables
If you are using a table for visualizing your data, you can still use the predictive planning feature in SAC. You can select the time granularity for the predictive forecast, the algorithm, the set of historical data to be used as the base.
3. Predictive Planning using Smart Predict
Smart Predict is the most advanced predictive feature in SAC. There are three types of Smart Predict scenarios– Classification, Regression & Time-Series – provided by SAC to address prediction on different business use cases. After selecting the scenario, you train the predictive model by feeding historical data from your data source. SAC uses an inbuilt machine learning algorithm that studies the input data and provides the best-predicted result set of various possibilities. Predicted data can be saved in SAC models that can be consumed in stories and applications.
Model Creation for Predictive scenarios
The prerequisite to run a predictive forecast in SAC is to enable planning options while creating a model, i.e., the planning model needs to be created to use predictive features. The data source can include flat files (CSV or excel) and acquired systems (BW, HANA). Live data source connection is not yet supported to enable the planning option.
We can agree that future prediction is directly dependent on the quality, granularity, and wide range of historical data. After you have completed data preparation, the next important step is to decide on the predictive scenario needed for each use case. Let us discuss each predictive method in detail in upcoming blogs.