Big Data, the next frontier for business intelligence where unimaginable amounts of unstructured content are mined for useful patterns, predictions and forecasts, is all the rage. But the reality is that Big Data is just coming onto the radar screens of CIOs and some see more immediate value in it than others, according to six members of the Enterprise CIO Forum (ECF) Council. The CIOs interviewed are General Dynamics Information Technology CIO and EVP Spain “Woody” Hall; Compass Talent Management Group LLC CEO and retired CIO Joel Dobbs; Campbell Soup Co. CIO Joseph Spagnoletti; China UnionPay EVP Chai Hongfeng; Deloitte Germany CIO Dietmar Schlößer; and Japan Tobacco VP of IT Hisayuki Hikichi. The interviews were conducted by ECF editorial director Bill Laberis. The Q&A was written by ECF community manager John Dodge.
ECF: Big data is loosely defined as that fast-growing data set of largely unstructured data (PDFs, video, Excel docs, etc., much of which is the byproduct of social media activities today) that organizations need or want to analyze but generally lack the infrastructure to do. Is big data showing up on your radar screen and if so, how important is it?
Hall: Big data, as defined above, is not yet a major concern within our enterprise though we are seeing some interest from our customers. Internally, we don’t support much video yet due to bandwidth cost issues and lack of business case. PDFs and Excel docs tend to be handled as individual artifacts and we haven’t seen much interest in treating them as enterprise assets. Clearly there is a knowledge management angle to work there, but we’ve got more urgent business needs on our plate.
Dobbs: I think that the relative importance depends a lot on the nature of the business a company is in. In the pharmaceutical industry where I spent most of my career and all of my years as a CIO, the availability of data on usage patterns and prescribing habits is critical to understanding the marketplace and targeting the marketing message. Social media also provides a wealth of data; however FDA regulations severely limit what companies can do in this space. Also, increasingly, states are limiting the ability of pharmaceutical companies to obtain and use prescribing data even when the data contains no identifiers. We should expect similar concerns in other industries as privacy becomes a greater concern.
One of the greatest potential uses is of course internally-generated information, especially in large companies. A by-product of the automation of business processes and the storage of documents on line is a trove of data that can be mined to generate everything from potential product ideas to forming communities of individuals with complimentary talents and interests (something we were experimenting with at my last company).
Finally, mining data on the Internet is a growing opportunity. IBM has some interesting technology in this space. At my last company we partnered with them to mine public data to try to understand how to construct more effective distribution systems for our products in third world countries. Interesting stuff indeed!
Spagnoletti: Yes, it is on our radar and a growing opportunity. It will form a major pillar in our development of a three-year IT strategic plan.
Chai: Yes, big data always shows up on my radar screen. Big data analysis transforms the human and emotional judgments into quantity analysis and a correct understanding [of the data]. Visits to this data could bring huge benefits to us. Big data is a reliable means to obtain a competitive advantage.
Schlößer: Big data is definitely showing up on our radar screen. I expect it to become much more important to us in the next couple of years. To me, the lack of infrastructure is not the key challenge though. What most organizations lack is a clear strategy on what exactly they intend to achieve with big data.
Hikichi: Unless clear definitions or concepts are streamlined for general business use and/or long forgotten keyword such as `AI` kind of solution technology is introduced, I would rather plan and invest more in balanced, structured, and response time requirement for `Big and Volume Data Warehouse` area instead of `Big Data` so-called unstructured data.
ECF: In general, what is your take on the importance or challenge of being able to analyze unstructured and semi-structured data in your organization?
Hall: We do see some value in being able to more efficiently locate and retrieve unstructured data from our collaboration environments. This internal need is not high on our priority list currently.
Dobbs: In my last company, this was huge for the reasons noted above. It is now a competitive necessity in the pharmaceutical business.
Spagnoletti: It is not well-integrated or easy to use in its current form. It does not lead to developing insights.
Chai: Data is a standard to directly measure the working performance, but data analysis does not simply summarize massive amounts of data. Idle data is useless and will only bring benefits to our business through processing and analysis. Therefore, how to filter the data is valuable and very important for us.
Schlößer : First, define and agree what you really want. Then, plan and build the infrastructure.
Hikichi: Big data definition is still vague in its contents, and mainly used by the vendor tools’ promotions, to date. For the enterprise, structured data volume and capacity management are categorized at a higher priority.
If it were called big data, then it is very important, [but] social media related data is not yet considered in mature phase, except for the marketing and promotion related strategies.
ECF: What steps or moves do you feel organizations should take today to deal with the challenge of big data?
Hall: It depends on the organization’s needs. The basic requirement would be to understand what the scope and value of unstructured “big data” in the enterprise might be and what value/investment makes sense in better managing it.
Dobbs: First, understand why. What information exists in today’s world? How can you get at it and what questions could possibly be answered?
The second question is, how? Do you partner with a third party or attempt to do all of this internally? I believe that partnering will increasingly make more sense.
Spagnoletti: At a very high level, define the purpose of the data, determine how it will be used for that purpose and create meaningful views that provide insight.
Chai: CIOs need to fundamentally reconsider their methods for processing data and reformulate an information management plan. The enterprise needs to offer the CIOs or other managers who need to visit the big data resources an opportunity to change the way the enterprise is using this information. IT managers must convince their business partners to face the challenges together and to ensure a certain level of control and collaboration. Otherwise [Big Data] may cause risks, increase costs and generate more silos.
Schlößer : Create a vision – what benefits do we expect? Agree on a strategy and plan and build the infrastructure.
Hikichi: Only actions we will be taking until buzzword and introduction phase cools off are objective review and benchmarking analysis.
For more information, see "Storing, capturing and analyzing information more effectively" and the "Advancing information management white paper."