The background of Predictive Analytics
Organisations are under constant pressure in today’s evolving business environment to understand and anticipate future business changes and predict the answer to the question “What will happen?”
However, many organisations, from an analytics perspective, are only able to answer the question of “What happened?” This is particularly true for those organisations that use the SAP business analytics toolset without any specialised non-SAP statistical and predictive modelling tools such as SAS or IBM SPSS. Predictive analysis is a concept that comes further up the analytics/maturity chain (see Figure 1). It is however becoming an ever increasing and important capability in the area of business analytics according to a number of research bodies. This is particularly so with the advent of real time and big data concepts within organisations and industries.
For those companies that do use specialised statistic and data mining applications, they are typically owned and controlled by data scientists who produce numbers which are seldom shared or visualised across the business. They are seen as expert user tools that don’t allow for instant predictive capabilities on the desktops of the users’ machines. SAP Predictive analysis is unique in this sense because whilst being a predictive analytics tool, its use is out in the business and this is where SAP feels the main user will be (see Figure 2).
So what is SAP Predictive Analytics?
SAP Predictive Analysis is complete data discovery, visualization, and predictive analytics solution designed to extend an organisations’ current analytical capability. It allows organisations’ business users to build predictive models to discover hidden insights and relationships in their data, from which they can make predictions about future events.
The application, which was made generally available in the last quarter of 2012, has allowed SAP to move quickly into the predictive analytics domain as a key vendor. It also uses the same code from the SAP Visual Intelligence application, allowing for rich and flexible visual representation of data as well as the capability to source data from a wide variety of sources and the functionality to enrich, merge and manipulate the dataset where required.
SAP Predictive Analysis is a client based application which offers a range of predictive analysis algorithm. It can be used as a standalone application or can be used in conjunction with SAP HANA, and depending on the deployed option the varying algorithm libraries are used (see Figure 3 for further detail). It’s the latter option of using Predictive Analysis on HANA which makes the use of predictive analysis very powerful, allowing for real time, in-memory processing of large historic datasets to predict future outcomes and actions.
The process of using the application is simple and initiative, whilst still using powerful and complex algorithms to empower the business users to derive meaningful and contextual results to drive effective business decisions and actions.
The application’s interface allows a business user to perform 3 simple steps to acquire, enrich and perform predictive (see Figure 4).
Predictive models can be saved and shared, with the results publishing to a file, database, SAP Streamwork or SAP BusinessObjects Explorer (desktop or mobile).
So what predictive algorithms can be used?
SAP Predictive Analysis uses a blend of native and R algorithms, which cover various analyses such as:
Time based predictions and forecasting
- Association/affinity analysis
- Clustering & Classifying
- Detecting outlying values
R is an integrated suite of software facilities for data manipulation, calculation and graphical display. It is open source software that is well trusted and recognised worldwide. More information about R can be found at http://www.r-project.org
One of the benefits of SAP Predictive Analysis is that the user works with the intuitive interface to access the complex predictive algorithms, without the requirement to explicitly understand the inner working.
Currently there are around key 15 algorithms if you use the application with non-HANA sources (processing on the client) and 10 if you do use HANA as a source (and therefore the processing is done in HANA). SAP are constantly increasing the algorithms available with each new service pack.
What’s the business value?
The main benefits of the SAP Predictive Analysis tool are that it allows users to discover and predict results in a matter of minutes without the need to involve IT or data scientists. Do you ever have the requirement to perform the following tasks, but are restricted by the applications you have available or the process you need to go through?
- Predict market trends and customer requirements
- Predict how price volatility will effect costs and demands
- See how changes in material supply will affect the entire supply chain
- Understand influencing factors in a sale, buying or manufacturing cycle
- Provide exactly the right offers and service levels to every customer
- Predict cash flows to manage your business more efficiently
- Understand customer sentiment and churn
SAP Predictive Analysis allows you to quickly perform these analyses which will ultimately drive your ability to predict and solve difficult industry scenarios and gain competitive advantage.
The application works in your existing environment, requires minimal learning or setup to be up and running and empowers business users with an intuitive application. The key three value points of the application are empowering the business, deriving actionable insights and meaning results. These are described in further details in Figure 5.
SAP Predictive Analysis bridges the gap between traditional “What happened?” analytics and advanced analytics, answering the “What will happen?” This will allow organisations to move quickly and cost effectively up the analytical value curve as identified in Figure 1. The product is competitively priced and integrates into your existing landscape, as well as integrating with the BusinessObjects BI platform.
The application is extremely intuitive meaning you can be up and running in a very short space of time and the technology is completely non-disruptive. The tool allows a business user to deploy complex predictive models with ease without the need for IT or data scientist skills or resources. Gaining key insights into future events and outcomes can help lead towards gaining a competitive advantage and moving from a reactive to a proactive approach.
Find out more about our services in Predictive Analytics and see latest video on Decision Trees here
To learn more about SAP Predictive Analysis see a demo and understand how it can add value to your business please get in contact at firstname.lastname@example.org