Big Data, it seems, has reached the main stream. Everyone’s talking about it and everyone’s wondering whether they should do it, according to new research from IBM. But it remains to be seen whether businesses and business decision makers really understand what Big Data actually is, or how to develop a strategy get the most of it.

IBM Big Data @ Work Survey

The findings of this research entitled "Analytics: The real-world use of big data" were based on the IBM Big Data @ Work Survey that was carried out in the middle of last year in 95 countries across 26 industries and which surveyed 1144 professionals working around big data.

The first thing it found was for many, even the term big data was confusing largely because it is used in association with so many different IT disciplines. It was used by respondents in the survey to describe all kinds of different things including: large quantities of data, social media analytics, advanced data management capabilities and a lot more.

However, it also showed that a number of pioneering enterprises are already starting to achieve breakthrough business outcomes.

While there are many ways of defining big data analytics, for the sake of this study IBM defined it as the ability to extract new insights from existing and newly available internal sources of information. It also refers to the technologies used to extract meaning.

IBM_Big data customer_centric outcomes.jpg


While the study goes through many different aspects of big data, like the four elements of big data -- volume, variety, velocity and veracity -- it also offers five principal recommendations for organizations to progress their big data strategy and maximize the business value from enterprise data. Those five recommendations include:

  1. Start with customer-centric outcomes
  2. Develop a strategy for the entire enterprise
  3. Start with data that is already available in the enterprise
  4. Identify business priorities and build the strategy on that
  5. Develop business case based on measurable outcomes

1. Customer-centric Outcomes

It is absolutely essential, the research says, that enterprises focus their big data strategies on efforts that will provide most business value. For most, this means starting an analytics strategy with customer analytics in order to provide customers with better services, which in turn should lead to better customer retention.

While this may seem obvious, the difficulties in doing this are growing every year as individuals become digitized and better informed of their choices.

The result is that enterprises will need to understand their customers as individuals, and need to invest in new technologies and advanced analytics to do this.

The ultimate goal here is not just to get to know the customer, but also to connect in a way that the customer sees as useful, which may come in the shape of more timely, informed or relevant interactions. Big data analytics, for its part, offers insights from big data that are increasingly important in these kinds of relationships.

IBM_big data sources.jpg

2. Enterprise-wide Big Data Blueprint

The blueprint of your big data strategy must cover the overall vision, the strategy and the requirements not on a departmental basis, but on an enterprise basis. This should result in the creation of a common, enterprise-wide understanding of how the enterprise intends to use big data to improve business objectives.

With it the enterprise will be able to identify key business challenges to be overcome, the business process requirements that define how big data will be used, as well as the architecture, data, tools and hardware needed to make the blueprint a reality.

It also provides the basis for developing a roadmap to guide the organization through practical approaches to the development and implementation of its big data strategy.

3. Start With Existing Data

To achieve short-term results as the big data implementation starts and gathers steam, enterprises need to be realistic about what they can achieve initially. For those that have implemented a successful strategy already that is providing business value, the easiest place to gather insights is from information that is already in the enterprise.

Doing this enables an enterprise not only to use easily available data but also the skills and software that are already in place. This provides immediate benefits as they make the business case for extending the big data analytics to include more complex sources of information and types of information.

Most successful strategies have started analyzing existing repositories of information while scaling data warehouses to handler larger volumes of information for future insights.

4. Business Priorities, Skills Investments

As the market matures, businesses are being forced to choose between a growing number of analytics tools while, at the same time, having to deal with a critical shortage of analytics skills in both the US and Europe.

Big data success hinges on finding a way around this, which has seen IBM set up special analytics colleges in the US and Canada.

But for the moment, businesses will have to work in a market as it exists now and that means investing in tools and skills. As part of this process the research suggests that new career models will emerge for individuals with the requisite balance of analytical, functional and IT skills.

For those that already have the skills in-house, enterprise must focus on professional development and clear career progressions; investment in these people at the moment should be a top priority for executives.

IBM_big data implementation.jpg

5. Measurable Outcomes

To develop a viable big data strategy and to ensure that there will be ongoing interest and investment from decision makers, enterprises need to ensure that the case for ongoing investment is based on quantifiable business outcomes. In other words, business leaders need to be able to see the advantages.

Businesses can do this by ensuring that there is active involvement and sponsorship from one or more business leaders when the original strategy is being developed and when the first implementations take place. Also of crucial importance here is ongoing cooperation between the business and IT departments. This should ensure that the business value of all the investments in big data analytics is properly understood.

This paper and the research around it have a lot of insights for businesses that are considering their first big data deployments. Like all other IT elements in a business it requiems thorough planning before execution. This paper makes a good starting point.