Predictive Analytics for Non-Stats Majors

At the Business Intelligence booth at SAPPHIRE NOW in Madrid 2012 (Photo: SAP)

The biggest announcement that didn’t happen at SAPPHIRE NOW in Madrid? A soon-to-be-released predictive analytics solution for SAP’s BusinessObjects BI portfolio. Luckily, visitors at the event could see the planned solution in action on the show floor and learn about its capabilities. Predictive Analysis, as the product is referred to now, allows business users in many lines of business and industries to conduct forecasts for revenue, sales, costs, and other factors. While this analytics solution is geared towards the “non stats major”, SAP also offers predictive capabilities for developers and hardcore data scientists on the SAP HANA platform. Importantly, the upcoming release will enable businesses to integrate Predictive Analysis with SAP HANA, thus addressing both audiences. Depending on the deployment, Predictive Analysis will then be able to access data in both SAP and non-SAP systems, including CRM and ERP,  universes in BusinessObjects BI applications, SAP HANA, SAP Sybase IQ server, Oracle, as well as from unstructured sources like social media.

Potential use cases for predictive analytics

The telecommunications sector presents one use case in the area of “turn management”. Using Predictive Analysis, companies could identify which customers are likely to end their contracts, and react accordingly. A company in the manufacturing industry might use predictive analytics to determine how to run its machines most efficiently. And banks could offer micro loans and credit to targeted customers.

While predictive analytics offers promising new capabilities in a variety of industries – and is the first step towards artificial intelligence in the workplace, say some – solution experts are careful to add that it’s not a replacement for human understanding of the business. “Just imagine,” said one solution expert during the demo, “if you conducted a sales forecast for a particular week in Madrid, and you didn’t take the general strike into account.” Outside factors like these, might not always be represented in the data sources being used.

But even if the analysis is based on accurate information, there are still risks associated with how users interpret that information. Especially business users with little to no analytic experience. Today 10% of people in an organization have access to analytics. Gartner expects that number to increase to 75% by 2020. It’s true that you don’t have to be a statistician to use analytic applications like Visual Intelligence and Predictive Analysis. With handy drop-down menus and clear graphics, they’re made for business users, not IT. But it will still be crucial for companies to educate their employees in this area in the coming years.

Next page: More new analytic applications from SAP

SAP’s predictive analytics solution is not yet generally available, but the company has announced support pack 5 for the SAP HANA platform (SPS5), which introduced new predictive algorithms for classification, clustering, outlier detection, and data preparation. These functions are more for the technical developer and experienced data scientist, who then make the results available for modeling and consumption by the business user. SPS5 will also support predictive model markup language (PMML) for the integration and sharing of predictive models.

Analytics for business users and developers

In addition to the update for the SAP HANA platform, SAP has recently released several other analytic applications. One of them, SAP BusinessObjects Design Studio, is also in SAP’s suite of BI products. This product introduces a premium alternative to BEx Web Application Deisgner, which is a tool that allows users to create HTML pages from content in SAP Business Warehouse. One of the main differences between the two products is that SAP BusinessObjects Design Studio is mobile, while the BEx Web Application Designer is desktop-based. In addition, Design Studio is multi-dimensional, allowing users to drill down into BW and HANA cubes and bring up only the data they want to display. It is based on the Eclipse developer environment and uses familiar standards like HTML5 and Java.

Compared to Design Studio, which is a decidedly developer-targeted release, the new batch of HANA-based applications address the business. Here is an overview of the latest releases. See the press release for more information on the usage scenarios and application details.

  • SAP Liquidity Risk Management application
  • SAP Accelerated Trade Promotion Planning application
  • SAP POS Data Management application
  • SAP Customer Usage Analytics analytic application
  • SAP Demand Signal Management application
  • SAP Operational Process Intelligence software