Big Data Analytics
The first steps toward achieving a lasting competitive edge with Big Data analytics.
We've all heard the phrase "Big Data is coming to the enterprise" many times over the past year. It is, however, something of a misnomer. Big Data has been with the enterprise for some time - what's new is the variety of means to deal with it.
The chief driver behind all of the new "Big Data" technologies these days is the belief that somewhere in the growing reams of unstructured data lies the means to gain a competitive advantage. Tapping into this institutional knowledge is the key to new business models, more efficient processes and even new products themselves. As Donald Rumsfeld would put it, it's a way to access "the unknown knowns."
According to tech consultant Dion Hinchcliffe, Big Data does not provide a means to deal with just more data. Rather, it incorporates analytics, analysis and management to turn data that is bigger, faster and more complicated into usable knowledge. In essence, you needed to compile what is essentially fragmented pieces of information into a cohesive, end-to-end data environment that is at once easy to access and yet represents a comprehensive view of an organization's collective experience.
That's one of the key differences between traditional data warehousing and modern Big Data analytics, echoes HP's Srinivasan Sundara Rajan. While warehousing works best in environments where data relationships are well understood - that is, structured - today's analytics need to keep pace with the hectic world of emails and social media, providing near real-time analysis and, ideally, keener insight into what is happening right now.
Naturally, it will take an enormous amount of IT resources to conduct that level of analysis on an on-going basis. Fortunately, IT-on-demand has become a reality with the advent of the cloud. Companies like Quantivo are quickly ramping up cloud-based architecture capable of handling Big Data analysis requirements. Quantivo's platform features tools that scan data for "monetizable patterns," such as online behavior and customer experiences, which can then be used to craft optimized processes and procedures. The company says it can handle billions of records spread out across multiple data sources with unlimited scale in both processing and storage. In short, no more compromises between top-notch performance and cost containment.
Still there is a danger in all of this, says former Gartner analyst Amrit Williams. Do you really want to build efficient and effective solutions for your data, or are you merely interested in chasing the latest trends? If it's the former, it would be wise to do a little pre-analysis of your own to see if your unstructured data actually holds the knowledge you seek. If so, do you then have the in-house means to make use of it, or for that matter, to even manage the Big Data platform needed to extract it?