Nomad_Soul - Fotolia
Big data is a journey, says Forrester analyst Brian Hopkins, and as with any other technological journey, companies planning to trek toward architecting for big data would do well to assess upfront just what they're looking to accomplish with their particular data sets. For many enterprises, this assessment likely starts with pinning down the increasingly murky definition of big data itself, a blanket term bandied about by marketing, technology and business departments alike. This first step is crucial, many analysts argue, toward establishing the connection between any big data investment and the resulting business outcome.
In this recap of SearchCIO's recent #CIOChat on architecting for big data, experts Tom Doub and Wayne Easterwood, respectively CEO and CIO of Centerstone Research Institute, discussed with fellow Twitter participants the various considerations associated with understanding and investing in big data.
What should CIOs consider before making a big data investment?
When it comes to big data purchasing decisions, businesses can't stop at semantics. Pinpointing exactly what big data means entails an organization looking at its business problem from many angles, including possible market differentiation opportunities, cultural issues, specific use cases and more, as Doub and frequent #CIOChat-er Tim Crawford pointed out:
#CIOchat A3: 1) Intended use of data 2) Volume of data processing required 3) Suitability for analysis 4) cost/reliability 5) scalability— Tom Doub (@tomdoub) September 24, 2014
Stephen Laster, chief digital officer for McGraw-Hill Education, can attest to the benefits of using specific data to drive his organization's big data technology investments. Before his team drafts a technology roadmap, "We decide, No. 1, 'What are we trying to accomplish to advance teaching and learning?" he told senior news writer Nicole Laskowski in a recent SearchCIO story. Only after they've answered that question do they move forward with deciding whether to build or buy each big data capability.
Especially in industries that deal with more sensitive data, like patient health information, the regulatory compliance implications of big data are particularly important, tweeted Doub and SearchCompliance editor Ben Cole:
When examining any of these factors, CIOs and other buying teams ought to view big data technologies as long-term investments, encouraged participant Chris Petersen. Fellow chatter Gloria J. Miller took it a step further, promoting Agile techniques to tackle technologies' pace of change:
Rick Sherman, a business intelligence expert and founder of consultancy Athena IT Solutions, is another proponent of using an Agile BI approach.
"Agile BI techniques can significantly reduce the likelihood that IT will develop something the business cannot use. In addition, it's very common for business users to refine requirements or identify new ones that they forgot about or assumed that the IT staff knew about," he wrote on our sister site SearchBusinessAnalytics.
What about big data analytics in the cloud?
Because any one product vendor's technology stack isn't as "complete" as it advertises, a single stack likely can't handle diverse, complex big data requirements, according to Miller, who brought up the strategy of turning to cloud providers for big data analytics:
Cloud tools, however, have yet to catch on in many organizations that are analyzing big data. One reason is their processing drawbacks, as editorial director Christina Torode and executive editor Linda Tucci note:
Easterwood also warns organizations to be careful with personal health information (PHI), particularly because they're protected by U.S. Health Insurance Portability and Accountability Act business associate agreements (HIPAA BAA):
And of course, no cloud discussion would be complete without debating its security, not just because many organizations are leery of sending mission-critical analytics data and applications into the cloud, but also because of the continuous spate of high-profile data breaches that dominate today's headlines:
What factors did you take into consideration before making a significant big data investment? Sound off in the comments section below.
See our first tweet chat recap for guidance on how to parse through various big data options. Then, head over to SearchBusinessAnalytics to see how some organizations found the best big data fit for their business. Lastly, check out Data Mill for case studies of companies deciding between building and buying their big data architecture.