OpTier Study Shows Main Issue with Big Data Analytics is Lack of Context

A new report has come out today from OpTier (News - Alert), which takes a look at big data Analytics and highlights key findings the company’s research study has made.

The initiative for the study was to better understand how companies are currently analyzing big data--in effect to find what the companies are lacking, and how they can take better advantage of big data. 


Image via Shutterstock

The study was led by Jacques Takou Tuh, an MBA student at the Kellogg School of Management , along with Adam Kanouse, CIO of OpTier. Together they conducted primary research with various business executives at many Global 250 companies, including one-on-one interviews, focus groups, and Q&A’s regarding the financial services, healthcare, media and entertainment, retail and telecommunications industries.

The results of the study were released today in summarized form in a research paper entitled “Making Big Data Analytics Fast and Easy.”

One of the main findings was that despite the hype surrounding big data analytics, few companies are actually able to apply one to the other. Application of analytics to big data for competitive advantage seems to be something most companies struggle with, due to the time and costs associated with the traditional approaches to analytics.

The study found that the majority of big data resides within a company’s data center, and not through social media vehicles as is popularly thought.

Most companies are also failing to gain advantage from their big data, in large part due to its distribution across too many silos. Big data is lacking the context and uniformity that is necessary to allow analysts to quickly leverage it, in other words.

Furthermore, the Data Preparation phase of Analytics accounts for the bulk (30 to 60 percent) of the time spent in analytics, as data is commonly saved without context.

These problems go along with the finding that among the companies surveyed, nearly all agree that context  would dramatically accelerate analytics, as it would allow a 50 to 90 percent reduction in time and costs spent in this respect.

But how much time is being spent exactly? According to the study, companies reported spending as much as two to three years setting up data warehouses, and two to three months on data set incorporation.

Other aspects of big data analytics which were found to be time-consuming were traditional statistical modeling of data relationships, and re-writing applications or re-building from the ground up.

“Based on the research conducted, a common theme that emerged was the need for a faster way to get meaningful business value--such as the interrelationships between various data sets--out of their Big Data,” summarized Russell Walker, a professor at the Kellogg School of Management. “Contextual data is key to drive business growth.”