(By Mani) EMC Corp. (NYSE: EMC) said last month it could pull Big Data assets together from EMC and VMware, Inc. (NYSE: VMW) to create the Pivotal Initiative as EMC is looking to recreate its VMware success by establishing a focused Big Data unit that it hopes becomes a standard platform.
Gartner's estimates for Enterprise Software include the Big Data infrastructure elements and imply a $100 billion market by 2015. Since the Pivotal Initiative is unlikely to get into the database market, Pivotal's total available market in 2015 looks more like $70 billion.
[Related -EMC Corporation (EMC): Margin Improvement On The Cards As Portfolio Expands]
On Dec. 4, EMC blogged that it is pulling together its Big Data efforts under Paul Maritz under the name the Pivotal Initiative to cover data science (analytics), technology (Hardware and software), and consulting.
The group includes employees and resources currently working within EMC's Greenplum and Pivotal Labs organizations.
The Pivotal Initiative is looking to build a complete and integrated Big Data ecosystem that combines the above with vertical industry specialization, as well as EMC's own scale-out x86 hardware Isilon.
Meanwhile, EMC needs to address a larger part of the IT budget as International Business Machines Corp. (NYSE: IBM) and Oracle Corp. (NASDAQ: ORCL) do. To move beyond its storage silo as it did with VMware and RSA, EMC is looking to build an infrastructure for Big Data application development, deployment, and operation.
[Related -EMC Corporation (EMC): Are You Ready For The Next Internet Boom?]
"We expect further description of this new business later this quarter, but it is a clear attempt by EMC to address a greater portion of the IT budget. It is possible EMC may spin-out part of Pivotal to create a currency as it did with VMware," UBS analyst Steven Milunovich wrote in a note to clients.
EMC views the current business intelligence/data warehouse versus Big Data spending split as 90/10 while the amount of data is 10/90. Spending should shift toward Big Data.
EMC hopes to become the standard platform for Big Data as VMware has for virtualization, which is a tall order. Scott Yara, a founder of Greenplum argues that traditional data warehouse vendors like IBM, Oracle, and Teradata Corp. (NYSE: TDC) will have some success re-platforming their customers. However, the unstructured market has different requirements that require a new approach.
He thinks Greenplum has an advantage due to a Hadoop-first mentality with its own distribution, SQL processing services, and better economics, particularly compared to Teradata's services-heavy approach.
"Although Greenplum today sells a $1mn appliance, it sounds like Yara expects most revenue to be software-based in the future," Milunovich said.
EMC's Pivotal Initiative has hardware and software for the run-time infrastructure, but the more significant spend will be on development. IDC's Digital Universe study found that less than 1 percent of the world's data is being analyzed today.
The actual growth Big Data experiences over the next three years will be highly dependent on overcoming the current enterprise trend of merely piloting Big Data.
An IBM survey supports the reports that Big Data projects are mostly in the pilot phase. While interest is high, with nearly two-thirds of enterprises surveyed use big data-like data analyses to create competitive advantage, concerns over delivery of measurable business value seem to be holding up projects.
IBM's survey lists a number of obstacles specific to Big Data projects, including understanding how to use big data, data quality, and a lack of analytics and technical skills.
Looking at EMC's success with VMware, the obvious question is, "Can EMC do it again?"
"Given the early stage of Big Data efforts in enterprises, the reputation EMC has with its customers, and the need for storage as a key element of Big Data, we believe EMC has a chance though the Big Data field will be highly competitive," Milunovich said.
EMC will need to define Big Data as a new category requiring new approaches, so IBM, Oracle, and Teradata don't subsume the business.
What seem to be missing in the PI portfolio are extract, transform, and load (ETL) capabilities, as well as more analytics expertise likely needed by industry verticals.
"With many current Big Data pilots spending up to 80% of their time on data preparation, we believe PI will be looking to acquire ETL capabilities sooner rather than later," Milunovich added.