IBM has the broadest and deepest Big Data product and services portfolio in the industry, as well as the market leading revenue to show for it. But IBM's greatest asset also lies at the heart of its biggest challenge. With such a diverse set of Big Data capabilities, IBM has struggled to unify them into distinct, compelling offerings. How IBM responds to the challenge of bringing together such a broad and deep set of technologies and services - many the result of $16 billion worth of analytics-related acquisitions since 2005 - into consumable and effective product offerings will largely determine the company's success (or failure) in the Big Data space and will have major implications for enterprise CIOs.
IBM has a number of positive factors going for it as it navigates the Big Data market. First, the company has a rich heritage of innovation stemming from its Research division, which is actively developing home-grown Big Data capabilities to compliment its acquired technologies (Watson, for example, is a result of internal IBM research & development). Second, IBM has years of experience integrating acquired technologies into its product portfolio. Third, the company has embraced, at least at the foundational level, the open source Big Data community, most notably in the adoption of Hadoop as the basis of its BigInsights platform.
On the negatives side, IBM is expensive. To date, IBM has positioned its services organization as a "wrapper" for its Big Data technologies. Tell us your business problem, IBM says to enterprise CIOs, and our consultants will work with you to integrate myriad technologies from IBM's Big Data portfolio to deliver the right solution. This approach is has proven extremely effective for large enterprises and IBM's government sector clients, but is cost-prohibitive for many mid-sized enterprises and puts pressure on CIOs to justify such large investments when there are competing, less expensive open source alternatives available on the market.
New IBM Big Data Offerings In Line with Market Demands
In terms of better packaging its various technologies into consumable offerings, the company took a step in the right direction last year with the debut of its PureData System family of appliances. Each appliance - PureData System for Analytics, PureData System for Operational Analytics, and PureData System for Transactions - is designed to speed deployment times and is targeted at specific workloads. IBM continued these efforts today with the announcement of a new Hadoop-based appliance.
IBM also announced number of enhancements to its InfoSphere BigInsights and Streams platforms, and new in-memory technology aimed at supercharging existing database technology.
The highlights of today's announcement include:
- BLU Acceleration: An internally developed technology that adds "speed of thought" performance capabilities to IBM's DB2 and Informix databases, BLU Acceleration leverages in-memory data processing, columnar architecture and data compression to provide near real-time query responsiveness. Initial use-case include interactive time-series data analysis via Informix and real-time reporting and dash boarding with Cognos business intelligence (BI) tools. IBM plans to make BLU Acceleration applicable to other data management products in the future.
- BigSQL: BigSQL is an ANSI SQL interface to provide BigInsights users a method for exploring and analyzing data stored in Hadoop leveraging existing SQL skills rather than requiring knowledge of MapReduce.
- PureData System for Hadoop: The latest PureData System appliance, PureData System for Hadoop bundles preconfigured hardware with the BigInsights platform, including advanced data visualization software and analytic application accelerators for specific use cases, all ready to use out of the box.
In order to meet the challenge of delivering compelling Big Data offerings, IBM must keep ahead of - or at least keep pace with - the fast moving Big Data market. Today's announcements indicate that it is indeed listening to the market. Specifically, feedback from the Wikibon community indicate there are currently three areas of particular interest to Big Data practitioners and enterprise CIOs. They are:
- Big Data integration with existing data management environments. Enterprises have invested heavily in data warehousing (DW) and BI technologies over the last two to three decades. While the goals of achieving a "single version of the truth" or a pristine enterprise data warehouse have largely gone unfulfilled, many enterprise have none-the-less come to rely on these DW and BI technologies to support mission-critical business operations. Enterprise CIOs are not going to rip-and-replace these technologies but are looking for Big Data solutions that compliment existing investments, can be integrated seamlessly into current IT environments, and provide new, game-changing analytics capabilities.
- Security, manageability and governance. Most of the innovation in the Big Data space over the last decade has been focused on new ways to analyze and otherwise monetize multi-structured, high velocity data. But before CIOs adopt a new technology to support mission-critical workloads, they need assurance that the solution is secure, manageable and able to conform to existing governance & compliance requirements. Big Data is no different.
- Accessibility. CIOs are less likely to invest in new analytics technologies that are only accessible to an enlightened minority within the enterprise than those that are leveragable by a wider population of business users. Feedback from the Wikibon community indicates that IT vendors should focus their efforts on developing Big Data platforms and technologies that are accessible not just to Data Scientists but to business users as well.
Today's announcements from IBM do not reflect unique approaches to Big Data storage, processing, or analytics. A number of vendors already employ in-memory storage to support real time analytic workloads, there are numerous Hadoop appliances currently on the market, and bringing SQL-like capabilities to Hadoop has been an ongoing theme in the Big Data community for months if not longer.
But they do indicate that IBM has its ear to the ground and is actively evolving its Big Data portfolio of offerings to meet the needs of practitioners and the enterprise. To wit, BLU Acceleration is aimed at increasing the value of existing DB2 and Informix deployments by providing targeted real-time capabilities where applicable. BigSQL, like Cloudera's Impala project and Greenplum's recent PivotalHD distribution, addresses the accessibility question by making Hadoop accessible via SQL, the most widely used data language in the enterprise. And PureData System for Hadoop is meant to decrease the time-to-insight by simplifying Hadoop deployments and by pre-integrating analytics and visualization software so users can begin exploring data soon after deployment.
Underlying each of these new and existing offerings is IBM's information governance platform. As mentioned, while analytics and "cool" new uses for data get most of the headlines, being able to manage and govern the complete Big Data lifecycle to ensure compliance with internal governance policies and external regulations is critical and should not be treated as an afterthought.
IBM Well Positioned, But Plenty of Hard Work Ahead
IBM, like all vendors in the Big Data space, still has work to do, however. Namely, while these and other recent announcements from IBM address many top-of-mind issues for Big Data practitioners, there is still a long way to go towards achieving the vision of a single, comprehensive Big Data platform as outlined by Wikibon and validated by feedback from its community of IT practitioners and business leaders. This will require further integration of IBM's acquired and internally-developed technologies, particularly innovative tools and technologies such as Watson, as well as its information governance and management portfolio.
Further, enterprises are still seeking killer use cases for Big Data analytics and applications, as we'll as guidance for transforming themselves into data-driven organizations.
IBM is well positioned, perhaps better than any other vendor, to address each of these challenges. The company is already doing so on the technology and product front, with today's announcements being the most recent examples. IBM is also providing important thought-leadership to the market, starting right at the top of the company with CEO Ginni Rometty. It should continue and accelerate these efforts.
Action Item: Enterprise CIOs and Big Data practitioners must balance near-term goals with long-term visions when it comes to investing in Big Data technologies and services. Press Big Data technology providers such as IBM to outline how they enable customers to achieve demonstrable, short-term ROI (aka "quick wins") while simultaneously laying the foundation for sustainable, flexible Big Data practices. Evaluate product portfolios based on their ability to integrate with and compliment existing data management technology, to provide value to power users and business users alike, and to apply information governance and security best practices.