How to Get Business Value From Big Data Analytics

By Samuel Greengard Print this article Print
Big Data Analytics Value

It's essential to adopt a broader scope about data and analytics, create a very flexible and agile IT framework, and build a strong foundation for data science.

89 Degrees relies on a Hadoop data storage framework, clouds, machine learning algorithms, APIs and a group of data scientists to push insight into new and sometimes unfamiliar territory. "The ability to process large volumes of data quickly is critical," Poultney says. "But we also require a great deal of flexibility so we can add new parameters and data points on the fly." In some cases, the company is dealing with daily data sets as large as 20 gigabytes for a single client, she notes.

Big data is more than a business initiative, and it's more than a collection of tools and technology.

"In today's business environment, it's necessary to implement and activate systems that support fast data," Poultney points out. "Big data isn't just a buzzword or an interesting concept. It's a way to unlock value. But to succeed, you have to identify the business problem you want to solve and determine how big data can achieve that."

Using Big Data Analytics to Gain Deep Insights

How can business and IT executives navigate the evolving big data landscape? How can they best put analytics and predictive analytics to use?

Clear Peak's Ducrocq says that it's important to recognize that IT is rapidly becoming a commodity, and organizations must focus on business value. This means that the entire organization must work together to spot value points. It also means that every organization must build a unique infrastructure that supports the value proposition and then continually evaluate the situation.

Hitachi Consulting's Bazzell advises organizations to think broadly and creatively about a data framework. This includes plugging in conventional data sources, as well as tapping into the internet of things and connected machines and sensors to produce both human and machine data analytics.

In fact, he points out that analytics now transcends human thinking. "If you look at the IoT more broadly as anything that produces data, receives feedback and adjusts behavior, then humans become 'sensors' as well," Bazzell points out.

Indeed, this emerging paradigm requires a new way of thinking about data and analytics that extends through the entire organization. At the same time, "Open the flood gates of data to data scientists, and let them do what they do best: explore and correlate seemingly unrelated data points," Bazzell recommends.

This doesn't mean giving analysts and data scientists carte blanche to chase unicorns and rainbows. It's crucial to "hold data scientists 'feet to the fire' so that they have the flexibility to explore, but they are accountable to the goals, timing and budget of the use case," Bazzell explains.

Make no mistake, a new era has emerged, and business and IT leaders must adjust and adapt.

"If a data initiative is driven by IT and the business side is not involved, you might as well shut it down because it isn't going to succeed," concludes Clear Peak' Ducrocq. "You have to understand the business problem or challenge you are attempting to solve, who cares and why they care, and what is the cost of doing nothing.

"IT must help the business understand whether the data challenge can be solved and, if that's possible, how to achieve maximum results."

This article was originally published on 2017-02-24
Samuel Greengard writes about business and technology for Baseline, CIO Insight and other publications. His most recent book is The Internet of Things (MIT Press, 2015).
eWeek eWeek

Have the latest technology news and resources emailed to you everyday.