If you type the phrase “Big Data” into a search engine today, the results will be immeasurable. There are literally hundreds of books, articles and websites devoted to the “Big Data” phenomenon, which three years ago didn’t exist. The concept of Big Data has gripped businesses of all shapes and sizes. I firmly believe that we’ve only begun to see the potential power of data, when used correctly, within the enterprise.
In late 2012, we at Tata Consultancy Services (TCS) decided to dig a bit deeper. We wanted to learn more about the current state of the Big Data marketplace and uncover insights on six core issues that we believed needed additional clarity. To that end, we launched a Big Data study, which surveyed 1,217 companies across nine countries and four regions around the world: United States, Europe, Asia Pacific and Latin America. The study was limited to large companies with revenues of more than $1 billion.
Knowing that the term Big Data is highly subjective, we allowed the survey respondents to determine whether or not their initiatives were in fact Big Data. We asked respondents to base their decision on the following definition: “The collection, processing and usage of large volumes of digitized data to improve how companies make important decisions and operate the business.”
With this definition, nearly half of the survey respondents (47 percent) dropped out — that is, they said they didn’t have a Big Data program at all in 2012 (though practically all of them said they will have one by 2015). It was with the remaining 53 percent — 643 companies in four regions around the world — that we moved ahead with the study. We collected findings from respondents via three channels: online surveys, interviews and secondary research.
The questions discussed in the study ranged from corporate investment in Big Data, Big Data activities, the types of data that are most important, how data is organized across functions, the biggest challenges to turning data into actionable insights and the future of Big Data.
I have to say, the results were interesting.
The first big finding – Big Data spending
The polarity in investments among companies from around the world was staggering. In 2012, 15 percent of companies surveyed invested at least $100 million apiece on Big Data initiatives — with more than seven percent investing $500 million or more. On the other hand, 24 percent of companies spent less than $2.5 million each.
This dichotomy reflects Big Data’s current position as an intriguing, yet immature, enterprise technology. Companies around the world are waiting — quite literally — with bated breath to understand the value of Big Data. They’re ready and willing to spend the money, provided that they can see the ROI of Big Data activation. In fact, our study found that by the year 2015 companies across the surveyed regions are expected to spend 75 percent more on Big Data with median spending expected to increase to 17.5 million (up from $10 million in 2012).
With such enormous investment, where’s the value?
Today, the three business functions that are directly related to revenue generation — marketing, sales and R&D — account for about 42 percent of total Big Data spending within companies while functions such as manufacturing, logistics, finance and HR account for only 30 percent of spending combined. One can assume that the budget is allocated to marketing, sales and R&D because they present an opportunity for the greatest immediate ROI, due to the nature of their business, e.g., real-time service; however, if you look long term, each and every business function has the potential to benefit from Big Data.
What is important to keep in context is that each function has its own needs to address, its own obstacles to overcome and, subsequently, its own ROI and Big Data benefits. This became increasingly clear when we explored the specific ROI numbers within each function. While marketing and sales had the biggest budgets they did not have the largest returns. In fact, logistics and finance expected the highest ROI on Big Data in 2012. Logistics managers expected a 78 percent return while finance managers expected 69 percent. Impressive numbers when compared to marketing’s 41 percent ROI expectation.
So, why does spend not always align with ROI? It’s because of the individuality of each business function. The marketing function sees Big Data as an opportunity to tailor campaigns and determine campaign and channel effectiveness, but in practice they often lack the technological savvy to handle the massive quantity and variety of the data.
In comparison, logistics managers see Big Data as an opportunity to monitor product shipments in real time, determine locations of inventory shrinkage and identify spikes in logistics costs and discern the where and why. Their ultimate goal is to cut down on waste, reduce over-ordering and minimize product loss. Logistics has specific, actionable goals in mind and they’re able to use Big Data to see tangible results.
While it delivers results, Big Data can also present big challenges
There has been extensive coverage about the challenges companies face in collecting, processing, analyzing and using Big Data in their business. Much of the discussion has focused on the “V’s” of data — velocity, volume and variety. Another challenge has been finding people equipped with the skills to handle the data — the data scientists.
We put these assumed challenges to the test to see what really was impacting our customers the most and found that technology didn’t top the list. Instead, the biggest issue that companies are facing is cultural: they are struggling to figure out how to force traditionally siloed business functions to break down their barriers, cross-collaborate and, together, achieve optimal results.
We believe that there are separate issues behind these organizational challenges that need to be addressed. First, how to get divisions and business functions to share data that they want to protect and second, how to knit together the disparate data from across the organization into one, easily accessible database.
From my point of view, this is an opportunity for the CIO to assert his influence and authority and be the great equalizer across business functions by formalizing a collaboration process and centralizing the data storage and analysis process.
There are several reasons why it makes sense to centralize the process. For one, it gives the data scientists on staff an unbiased look at the data across the business. From a recruiting standpoint, a central data hub is an attractive option for emerging data scientists. It gives them the opportunity to learn and grow in a collaborative environment. Among our survey respondents, we found that the companies that had centralized Big Data (about 37 percent) were far ahead in investing and implementation of data programs.
From the present landscape to the future horizon
Over the past few years, we’ve seen the cloud, mobile technologies and social platforms move into the enterprise. They’ve done so in silos and they’ve been effective in that approach. Data collection and analysis will never reach its full potential in a silo. And I expect that as companies begin to gain a better understanding of Big Data’s overall benefits for the business, we will see an increase in investments, corporate initiatives and job opportunities within the data space.
For companies in the IT and consulting industries, this poses a unique opportunity. We need to act as mentors and guides to our customers to help them better understand the Big Data opportunity and to develop the tools they need to be successful in their endeavors.
To learn more about the TCS Big Data study, visit: http://www.tcs.com/big-data-study/Pages/default.aspx
Dr. Satya Ramaswamy is global head of TCS Mobility and Next Gen Solutions Unit, the global organization within Tata Consultancy Services responsible for bringing benefits of mobility and distributed computing technologies to customers across different industry verticals. He has more than 18 years’ experience in various engineering, product management and strategy roles. His experience spans mobile phones, packet switching infrastructure and high-reliability mobile application infrastructure. He holds five U.S. patents with eight more pending.