Any enterprise manager, or CIO, who thinks they may be able to avoid Big Data solutions, or managing Big Data, should probably read the recent research from IBM, which shows that not only has Big Data analysis become a core part of business decision making, but that for 63% of enterprises surveyed, Big Data analysis is giving them a competitive edge.

2012 Big Data @ Work

The research, which is contained in the "2012 Big Data @ Work Study" is not just about IBM promoting the massive investments it has made in Big Data and analytics in recent years, but a genuine attempt to get behind the enterprise thinking on how and why they use Big Data analytics.

Carried out in conjunction with the University of Oxford in the UK, the report is the result of a survey of 1144 business and IT professionals across 95 countries, which found that Big Data is no longer confined to IT processes and IT departments. At this point in time and for many businesses, it is a core element in their day-to-day business.

Introducing the report, its authors go so far as to suggest that Big Data is even transforming society. However, for the sake of this overview, its principal purpose is the transformation of processes, organizations and entire enterprises.

Big Data Business Edge

The starting point here must be the 63% of enterprises that say Big Data is giving them a competitive edge. We use this as a starting point because in IBM’s 2010 New Intelligent Enterprise Global Executive Study -- cited by IBM -- the equivalent figure is 37%, up 70% in less than two years.

Big Data, the survey showed, clearly means many different things to many different people, but in all cases it appears to bring with it enhanced business performance. One thing that was notable was that it is not being used in any significant way with social media.

IBM Big Data Business Information.jpg

Given the headlines linking social media and Big Data, this is probably surprising to many. However, the research shows that only 7% of respondents are using it to analyze social data, and less than half of respondents said that they collected social media data for analyses. Instead, IBM says, the vast majority are using analytics and Big Data analytics on information that is already contained in enterprise repositories.

Big Data Problems

Big Data and the attempts to analyze it has been an ongoing problem for some time, the report says, but there are two emerging trends at the moment that are making current problems and related Big Data activities different than before:

  1. Digitization: All data and information is being digitized with the result that there is large amounts of real-time data across all industries and repositories, much of it non-standardized data.
  2. Advanced analytics: Current analytics offerings are able to extract meaning from data at a level, and with an accuracy, that would have been inconceivable in the past.

In response to this changing landscape, and with the emergence of these new trends, the research found that many enterprises are taking a pragmatic approach to the problem of Big Data.

Those that are dealing with it best are those that are adopting a business first approach by identifying business needs first, then satisfying those needs and tailoring them to infrastructure, data sources and analytics availability to support business opportunity.

There are also enterprises that are using the information that they possess already, as well as deriving information from newly available internal sources, to create a clearly defined Big Data strategy -- building up their infrastructure to deal with existing needs, or perceived future needs.

Big Data Definitions

To even begin to fulfill these needs, it is important that enterprises understand what exactly is meant by Big Data. It may seem obvious, but in the variety of elements that were mentioned, it seems clear that it means all kinds of things to all kinds of people.

This was clearly shown in the research. In the response the researchers received to the question asking participants to identify two characteristics of Big Data (see below), you can see a lot of different answers were provided:

IBM_Big Data Definitions.jpg
IBM Big Data Definitions

Enterprises were divided on whether Big Data is dominated by greater volumes of data, new types of data, or real-time analysis of data. In fact, the responses were divided into even more categories than those three.

Summing the responses up, the report says the most useful way they found of characterizing the four major components of Big Data was: Volume, Variety, Velocity and Veracity.

These can be described as:

  • Volume: Refers to the mass and quantities of data that enterprises use to try and arrive at better decision making.
  • Variety: The numerous different types of data and its sources.
  • Velocity: The speed at which data is created, processed and analyzed.
  • Veracity: The challenge of providing accurate and reliable data, and cleaning as far as possible the unpredictability of data.

It is the combination of these four characteristics that creates competitive advantage in the digitized marketplace.

And it seems that most organizations are looking at Big Data deployments, if they are not already in the early stages of Big Data efforts. The research shows that a large number of enterprises are:

  • Actively studying Big Data for future deployments (24%).
  • Defining a roadmap related to big data (47%).
  • Already deploying Big Data solutions (28%)

A further and more detailed analysis of the responses also demonstrated a number of emerging trends that are common to all verticals. They consist of:

  1. A focus on consumer-centric objectives.
  2. Extensible and scalable information management.
  3. Use of existing, internal sources of data for initial deployments.
  4. Need for advanced analytics capabilities that are often lacking.

Big Data and Business

It is with these trends in mind that IBM recommends enterprises build their Big Data strategy. By doing so, it says, enterprises will be able to solve the business challenges posed by an explosion in the amount of data available.

IBM Big Data Business Outcomes.jpg

It is not just about using the most advanced technology, IBM says, but combining technology planning around business objectives taking into account the changing nature of data and the potential business case for each of those assets. In light of that, IBM offers the following four recommendations:

1. Focus on the Customer

Big Data initiatives should start with useful business goals so enterprises should be looking at Big Data with customer analytics first, which will give them deeper customer enterprise insights for actions. Mass digitization means that customers expect more. The result is that enterprises will have to invest in new technologies to keep up with expectations, or get left behind.

2. Big Data Blueprint

This should contain an outline of an overall vision of where the company is going with Big Data, the strategy for deployment and use, and the business requirements across the enterprise. It also identifies the key business challenges, how Big Data will be used and the infrastructure and hardware needed to overcome problems.

3. Business Priorities

Enterprises need to identify the problems they are facing and deal with those problems through analytics technology and skills acquisition. This is particularly important in light of the shortage of skilled labor in the field of analytics. It is anticipated then that new roles and career models will emerge as a result of this. Cultivating existing skill-sets should be a priority.

4. Business Outcomes

To develop a comprehensive and viable Big Data strategy enterprises need to develop a solid, quantifiable business case around the acquisition of technologies. It is important, IBM says, to have active sponsorship from business executives throughout this process. Equally important is strong, ongoing collaboration between business and IT. The business case is being made based on the following anticipated advantages:

  • Smarter decisions: Analysis of new sources of data to improve quality of decision making
  • Faster decisions: Enables more real-time analysis to support decision making

This is only an overview of this research that dives into each of these issues in greater detail. However, the bottom line with this, as it is with any business-related IT problem, is that enterprises need to plan their deployments.

There is no point just throwing money and technology at a problem if hoped-for results are not clearly defined and deployment strategies clearly described. Those that do will end up having to backtrack in the future in an attempt to untangle an almighty IT mess.