You may not need Big Data after all | #MITIQ

WarningThe business buzzword over the past two years has been “Big Data”. Companies are trying to figure our how they can leverage their collected data and translate it into a competitive advantage. However, according to the Director of MIT’s Sloan School Center for Information Systems Research, Jeanne Ross, this approach is not necessarily a one-size-fits-all for today’s organizations.

Ross, co-author of the article ‘You May Not Need Big Data After All’, cautions businesses against buying into the hype around Big Data.

“I think you grow into Big Data,” Ross notes. She explains that there are companies who find the competitive advantage works within their specific industries. As an example, she notes that the oil and gas industry has long employed Big Data for helping them to decide when and where they should place a billion dollar well. The success in one industry, however, doesn’t necessarily translate into success in others. “Many times we know great things about our customers. We just haven’t figured out a way to address them.”

When asked if the fear is misplaced that some companies feel in that they can’t address the Big Data they have, Ross states, “No, not misplaced at all. If you don’t think you can do it, you probably can’t.” For organizations recognizing the potential value of Big Data for the first time, this news could be disheartening.

“I don’t think most companies are data-driven,” explains Ross. “I think they are metric driven.”

This differentiation is important. Today’s companies can respond to certain kinds of data but in order to truly be a data-driven organization, they have to recognize which data is important. As an example, Ross cites Foxtel, a pay TV service based out of Australia.

“They saw what products were going out and what channels people wanted,” she states. Even with that information they were unable to make strategic decisions. “They went back and started looking at segments and realized what ‘data driven’ would be. They didn’t have the stomach to go back and do that.”

Where the CDO fits in

Discussing the emerging role of the CDO, Ross explained that too often there is a propensity to assume that once a CDO is brought into an organization all data issues will be addressed by that role and that little to no further attention is required. With Gartner projecting a 25 percent adoption of a CDO role in companies by next year, Ross claims most companies likely don’t need to create this position.

The key to running a successful organization is identifying and maintaining a single source of truth with respect to data. Many divisions within a company will manipulate data to show that they are running at a profit or contributing significantly to the organization’s bottom line. In the long run, this can be detrimental to the company because different data can show different outcomes.

Once companies adopt a single source of truth in their data, Ross believes it is of utmost importance that it is adopted in a top-down strategy. “We have to let people know mistakes have to be made. The faster you make mistakes, the more you can learn and the faster you can grow.” This strategy is ineffectual, however, if you start in the middle of the organization as people will be less willing to admit mistakes and failure if it hasn’t been adopted into the company’s cultural model.

The swiftly moving current of technology, especially over the previous five years, should be viewed critically by companies hoping to somehow gain a competitive advantage. Leveraging Big Data requires more than just a willingness to throw money at the problem. It requires a full understanding on the part of the company as a whole.

(Originally published at

photo credit: Free Grunge Textures – via photopin cc

Making Medical Data Better Data

large_2956265250In just over a week, the team here at SiliconANGLE will be setting off for Cambridge, Massachusetts for the 7th Annual MIT Chief Data Officer and Information Quality Symposium held on the MIT campus. The theme this year is ‘Big Data Demands Good Data’. In preparation for the event, we will be presenting a synopsis of some of the important topics scheduled to be covered in conjunction with other academic and real-world applications that apply.

One of the first sessions, to be presented by Dr. David Levine, Vice President of Informatics/Medical Director of Comparative Data and Informatics for United Healthcare, will address the need for improving risk models based on predictive analytics.  According to the abstract of Dr. Levine’s presentation, he will discuss how the advent of and improvements in data collection can be utilized to drive improved performance in health administration and patient care.

Writing recently for FierceHealthIT, Dan Bowman discussed how a recent survey of hospital CIOs found big data in the medical field was severely lacking in its efficacy. Bowman cites a healthsystemCIO.comsurvey in which 76 percent of respondents claimed their vendors were too often over-promising and under-delivering with their big data solutions.

In the same survey, it was found a full 52 percent of respondents admitting to not using their big data applications at anything approaching any level of sophistication. This figure is supported by the fact two-thirds of those who took the survey claimed their organization had neither the manpower nor the skill to take advantage of analytical tools at a high level. Each of these factor into one CIO stating the big data market, as it pertains to healthcare, isn’t likely to reach maturity for at least a few more years.

The challenge, it appears, is for the healthcare field to better understand the mountains of data they collect and learn how to better pore over it to help in establishing better predictive models for patient care and hospital administration. According to Chris Belmont, CIO at New Orleans’ Ochsner Health System, “We have the data points. We just have to do a better job of getting our hands around the data and understanding it better.”

What is the real-world benefit of achieving an optimal understanding and usage of data in the medical field though? Jeff Kelly of discussed the necessity of improving the collection and analysis of big data for those fields represented in the Industrial Internet in a posting entitled ‘The Industrial Internet and Big Data Analytics: Opportunities and Challenges’. Shockingly, he stated as much as 43 percent of an estimated $2.75 trillion in healthcare spending (for 2012 alone) was applied to unnecessary procedures and administrative waste. That abhorrent figure will be significantly reduced as the Industrial Internet is increasingly utilized by hospital CIOs to target administrative inefficiencies, eliminate waste and improve patient outcomes.

Dr. Levine’s presentation seems to aim at addressing the better utilization of data for improved patient outcomes. His work with United Healthcare, in tandem with their membership organizations, is striving to make predictive models more meaningful and actionable with the aim toward driving improved performance.

photo credit: Funky64 ( via photopin cc

(Originally published at