The Big Game


The Big Data hype is encouraging information managers to consider substantial technology investments. But amidst all the marketing hype, how can enterprises actually derive significant business value and profitable scenarios from Big Data?

Gaming on Big Data

Organizations across industries are showing significant interest in the Big Data hype. CIOs are allocating substantial IT budgets to be able to derive greater value from their information infrastructure by exploiting Big Data. Enterprises, recognizing the value Big Data can help deliver in achieving better business results through valuable actionable insights, are prepared for increased IT spending to maximize investment outcomes from their initiatives.

Yet, amidst all the hype, interactions with substantial client segments reveal that few organizations have adequate strategies in place to effectively execute a Big Data initiative and apply the Big Data ecosystem. This is quite understandable. Big Data initiatives promise to enable businesses in achieving better decision-making by discovering hidden insights previously unattained, and automating some business processes that could otherwise not be automated.

The challenge facing enterprises though is: how to obtain these benefits? Enterprises must evaluate their existing information infrastructure, understand the data landscape, the associated sets of technologies, and develop the necessary new skills to uncover insights previously unraveled. From the enterprises’ perspective, understanding the differences in available technologies and how they interact is crucial to maximizing Big Data investment. While determining the right mix of existing technologies, CIOs will also need to focus on expanding their applications portfolio beyond traditional components – which requires new skills and analytical approaches.

Information managers must select and incorporate various existing and new technologies, and available data sources to maximize on the investment on Big Data. The critical challenge, however, is that these technologies exist in a complex, intertwined set of offerings from COTS (commercial-off-the-shelf) applications and open source providers. Moreover, the applicable set of technologies largely varies based on a multitude of use cases.

The core idea of Big Data is to conduct analysis on large volume and diverse types of untouched data to be able to derive insights for accurate decision making. Analyzing such massive scales of dark or unstructured data requires large storage capacity to accommodate large volumes of data, as well as advanced tools and methodologies to achieve key outcomes, e.g., predictive modeling.

Determining the correlation of technologies, people and processes within the organization and associated functions within its network, as well as outside, is also a critical feature of the Big Data ecosystem. As such, unlike previously encountered use cases, new scenarios require highly linked and connected data types as well as specialized set of technologies to extract valuable insights that remained previously uncovered. These untouched insights can bring tremendous value – but the opportunities do not come without great challenges.


Addressing the Big Challenge

As data sources increase in volume and data types diversify to become more complex, traditional business intelligence will fail to enable decision makers in deriving value from existing data to its fullest extent. With more complex data scenarios, information managers and decision makers will need to look beyond existing business intelligence for appropriate insights and move to advanced techniques such as predictive and prescriptive analytics to derive greater value beyond hindsight-oriented business intelligence.

Enterprises that have successfully implemented the Big Data infrastructure will now look to evolve the physical mainframe information infrastructure of their Big Data platforms to virtualized environments. The cost of maintaining physical data centers to accommodate such massive volumes of data will necessitate the evolution of the private cloud.

Here, with exponential growth in data, information security also becomes a major challenge. This will require defining and implementing stringent information governance and control mechanisms to manage data from various sources and for various use cases.

Data sources frequently include personal or sensitive information of users, customers etc., or proprietary information of internal and external identities. Such sensitive information runs the risks of possible mishandling and misuse. The information security factor would be paramount while addressing Big Data. As such, managing Big Data sources will require information managers to apply robust data stewardship methodologies to ensure appropriate acquisition, monitoring and usage of data from multiple sources.

Deriving value from Big Data may require several sources of data with varying levels of structure and relationships. To achieve multiple positive technical and operational outcomes, information managers will need to deploy several different sets of technologies in combination. This will also require information managers to evaluate existing information and architecture elements to ensure that they adequately support the increased volume of data and variety of data types.

Developing defined enterprise Big Data strategies for their organizations will adequately equip CIOs in linking several IT oriented strategy components with their organizational strategic goal of their Big Data initiative. Big Data strategies that are enterprise wide – beginning from business units of its origination and stretching across the organization – will enable information managers maximize value from their investment. In a world that is increasingly becoming digitally enabled, data will play a central role in making critical business decisions.

Data for Decisions, Information for Action

Big Data is not an entirely new phenomenon. High volume of data with varying attributes has existed for many years. Why more and more enterprises are talking about it now is because of the availability of affordable solutions and tools to efficiently process, manage and analyze this large data. By extracting valuable insights from large data that lay previously rested, enterprises are looking at gaining competitive advantage founded on critical information sources. The question that arises now is: how to turn information into insight and transform data into decisions?

Information is expanding and becoming a massive source of more strategic opportunities. The coming of new technology trends, bolstered by the confluence of mobile device technologies, cloud infrastructure and the integration of additional information from social networks is bringing significant sources of insights which, if effectively explored, can enable decision makers make better, fact-based decisions.

In a digitized world, where organizations are increasingly creating and storing more transactional data in digital form, advanced analytics can enable businesses to collect, analyze and act upon data in real or near-real-time processing capacity. Sophisticated analytics and automated algorithms can augment or potentially substitute human decision making1 by exploring and exploiting valuable Information sources and automating critical processes. As we progress, Big Data will be the breeding ground for the analytics application age where advanced techniques (e.g., new types of algorithms and analysis best-practices) will be scaled up to support new technologies (e.g., storage, computing, processing capacity and analytical software).

From Hindsight to Insight to Foresight

Amidst much hype, it is critical for CIOs to understand the essential implications of initiating Big Data and what the ultimate objective of implementing a Big Data initiative for their enterprise would be. The qualitative substances of most Big Data initiatives largely vary with each case. Big Data initiatives are not intended to produce highly precise answers to very specific questions based on data gathered for that purpose, but rather they are attempts to understand correlations between multiple sets of data that may have been gathered for unrelated purposes2. Analysis of data or information is based on probability and reliant on the degree of accuracy of inputs. Generally, increasing the accuracy of the prediction is not done by determining the accuracy of each individual data record, but by increasing the density of the data (e.g., by adding more data points) and optimizing the number of attributes. This does not mean that traditional data warehouse and MDM will be replaced with the coming of Big Data. In fact, it is critical to understand that it is traditional data warehousing and master data management that will provide the basis for information governance of the larger Big Data ecosystem. However, simple application of existing information governance practices to Big Data is unlikely to deliver results – for the existing governance of data may be insufficient, or be compliant to different sets of organizational objectives. As such, deriving valuable insights from existing data will require clinical information governance practices tailored to the organizational objective of applying the Big Data ecosystem and drawing value from the investment.

Make your Big Move

TekMindz’s dedicated Center of Excellence for Analytics has been consistently interacting with its enterprise clients to understand their critical information governance challenges and working with them to rationalize and integrate their legacy information infrastructure with Big Data. We expect that through year 2014 and beyond, enterprises will begin to really derive substantial value from data as analytics will scale up to improve critical business decisions. At TekMindz, we are developing critical capabilities to support the digitally driven information age.

More on Big Data & Analytics offering