reborn55 - Fotolia
Increased interest in predictive analytics and big data is prompting many organizations to reassess their SAP analytics tools -- not only from SAP itself, but also from third-party providers. While SAP is among the market-share leaders, vendors such as IBM Cognos, Qlik, Tableau Software and SAS have also worked long and hard to gain reputations as analytics specialists.
So when does it make sense to turn to an outside vendor for SAP analytics, and what are the challenges of integrating third-party tools with SAP data?
Organizations may have several reasons for considering third-party analytics. One reason is a belief that SAP analytics tools won't be as effective with non-SAP data sources, according to David Loshin, president of consulting firm Knowledge Integrity. Another reason is competitive differentiation. "Any organization with access to SAP's analytics tools can perform the same analyses, performed the same way," Loshin said. "However, if there are particular analytic applications that comprise your organization's special sauce, it may be [preferable] to execute those analyses using a tool suite that is not shared with all your competitors."
Other motivations include the need for a more diverse set of analytics models, for less complex tools for business analysts and, Loshin added, to analyze ERP information with data managed in other platforms such as a data warehouse or the streamed, unstructured data in a Hadoop data lake.
Those are most of the advantages third-party vendors claim to offer. "A lot of times customers are using Qlik because of the complexity, quite frankly, of getting information out of the system," said Michael Distler, Qlik's director of product marketing. Third-party SAP analytics vendors also tout their prebuilt "connectors" for retrieving data from particular SAP platforms or to perform common tasks such as reporting. "We have about eight or nine different ways to connect to SAP," Distler added.
The interface between the SAP and third-party worlds is typically Business Warehouse (BW) -- SAP's longtime data warehouse -- or the newer HANA in-memory database and analytics platform on which SAP's development efforts are intensely focused. "Quite often, companies have made a decision that Business Warehouse will be their data repository," Distler noted.
SAP ERP systems lack connectivity options, and HANA is becoming a solid integration platform for performing analytics on data stored in those systems, according to Robert Green, director of product management at Tableau Software, another provider of analytics products certified by SAP. HANA also helps Tableau users get their data faster compared to a traditional database, he said.
"We've had steady releases of new functionality that has kept up with the evolution of HANA," Green explained, citing how Tableau's software has remained compatible with changes in the way HANA handles parameters and variables in database tables. Much of Tableau's demand comes from SAP owners who are using a hybrid of the two integration platforms -- a product nicknamed BW on HANA -- for key SAP analytics applications such as financial reporting and operational performance management. "I see a lot of BW on HANA now," Green said.
SAS, a longtime major player in business intelligence (BI), is another partner selling SAP-certified products. "SAP has provided open interfaces to their systems, which SAS has leveraged," said Nancy Bremmer, SAS's senior manager of alliances and channels. The vendor's integrated architecture reduces data movement to the bare minimum, she said, adding, "It is moved very fast and only temporarily to take advantage of faster processing."
Not so fast: Can SAP analytics tools compete?
Third-party vendors claim to excel at predictive analytics and user self-service -- two leading-edge BI capabilities -- but SAP concedes nothing to them.
SAP has spent several years acquiring or developing analytics tools that have kept pace with the competition and now overshadow its standard BI front end, BusinessObjects. Besides HANA and its self-service visualization software, SAP Lumira, the company markets several alternative packages to third-party tools and has been making some of its technology readily available as software as a service (SaaS).
"SAP has actively sought to create a reasonable set of tools for analytics," Loshin said, singling out two products that comprise SAP Predictive Analytics. "The Automated Analytics product automates data preparation and predictive modeling and allows users to implement a broad range of modeling techniques, including classification, regression, segmentation and social-network analysis. The Expert Analytics product is targeted to professional data analysts and data scientists who are more familiar with hands-on analyses using predictive modeling and statistical programming."
As with SAP connectivity for third-party tools, much of SAP's advanced BI capabilities depend on HANA.
One major reason is HANA's ability to handle "hot, cold and warm" data in a dynamic tier structure that gives users flexibility as their needs evolve, said Jayne Landry, SAP's global vice president and general manager for BI. "Think of it as having tiering automatically across your data holdings," she explained. "If you've got a large amount of data you don't know the value of, HANA is a great data store. Previously, you had to make a decision about how much data you wanted to keep." With HANA dynamic tiering, organizations can instead keep lower-priority data in a "cold store," such as Hadoop, and move it into warm or hot storage as their needs change.
SAP finds itself in the typical position of technology vendors that have little choice but to provide respectable integration with their competitors' offerings, while striving to give their own products a leg up. In this regard, SAP's certification process for third-party analytics can be instructive. The first and most basic level of certification is connectivity. The second level is security. The third and most sophisticated level involves "code push-down" capabilities that allow complex calculations and queries to be "pushed down" from SAP's programming language, ABAP, to the database layer where they can run more efficiently, said Ken Tsai, SAP's head of cloud platform and data management.
The third level is also where prospective buyers will see most of the differences among analytics products, Tsai said. According to Tsai and Landry, SAP has already done much of the push-down work and incorporated it into Lumira and SAP Cloud for Analytics, a SaaS offering that's built on HANA.
Read a review of SAP Lumira
See David Loshin's buyers guide to big data analytics
Understand SAP BW on HANA