SAP may be pushing its new in-memory analytics appliance, called HANA, as the company’s response to the Era of Big Data, but processing and analyzing increasing data volumes is just one part of the story. The other, more important part for SAP, in my opinion, is speed.
For those unfamiliar with it, HANA is an analytic appliance that stores data in memory rather than on spinning disk, enabling near real-time analytics. Operational data is integrated as is created with historical transaction and analytic data. HANA can process both structured and unstructured data, including social networking and machine-generated data, so enterprises can analyze larger data sets – up to hundreds of terabytes — than can practicably be analyzed with traditional data warehousing technology.
But Big Data, as defined by Wikibon, HANA is not. As Wikibon CTO and premiere storage expert David Floyer defines it, Big Data involves petabytes, even exabytes of distributed, unstructured data with flat schemas and few complex interrelationships. From a data volume perspective alone, HANA does not fit this definition.
Nor do Big Data processing technologies, the most popular being open source Hadoop, allow for data analysis in real-time. HANA does and that is where SAP should focus its messaging.
Take SAP customer Lionsgate Entertainment Corp., for example. Leo Collins, Lionsgate’s executive vice president and CIO, told me his company has been using a preliminary version of HANA to explore terabytes of detailed, granular data, like how movies performed in theaters and related operational expenses. While the data volumes are large – multiple terabytes – they aren’t Big Data large.
The benefit of HANA Collins stressed to me was speed. HANA allows Lionsgate to perform the same type of analytics it was running before, but at a much faster rate. This means Lionsgate can perform more queries and more varied types of data analysis in shorter periods of time, leading to new insights and answers to “questions we never thought to ask” before, as Collins puts it.
John Furrier, Internet pioneer and founder of SiliconANGLE, calls this approach “Fast Data.” That’s a perfect description, in my opinion, and one that I think will appeal to many of SAP’s customers.
Companies currently exploring Big Data techniques like Hadoop are typically on the cutting edge. They are mostly young web companies and start-ups with cultures that promote risk taking and experimentation. In short, they are not typical SAP customers.
Most SAP customers are more traditional enterprises that have long relationships with the German software maker’s ERP platform. More then anything they want stable upgrade paths, maintenance and support to keep operations running smoothly. Many are also interested in running data analytic against increasing amounts of data stored in both SAP and non-SAP systems. But they do not, for the most part, employ teams of data scientists needed to experiment with still developing technologies like Hadoop for distributed computing and Big Data analytics.
For these companies, HANA is an excellent fit. It leverages innovative but proven in-memory technology to enable high volume, real-time analytics against mostly in-house corporate data. It gives SAP customers the ability to gain insights from their data quickly to make fast, fact-based business decisions.
HANA is all about Fast Data, not Big Data. But that’s not a bad thing. With over 170,000 customers, almost all of which want to make quicker, better business decisions to stay ahead of the competition, HANA puts SAP in a good position to expand its analytics business.