Executive Summary
- Business leaders and professionals are getting wary of big data, disappointed by poor results and difficulty to generate real value
- Big data data initiatives can fail for because of unclear strategies, but they mostly fail because of ineffective operating models
- To attain valuable business results from big data, organisations need to institute an operating model supported by right people, result-oriented process and appropriate technology
A Brief History of Big Data
The term big data first appeared in an article published by Michael Cox and David Ellsworth in 1997. Cox and Ellsworth used it to describe the visualisation challenges faced by IT infrastructure and computer systems in handling large sets of data. By the early 2000’s, the explosive growth in volume, velocity and variety of information had become synonymous with the term big data, but much of the discussion over this period of time was about infrastructure – how to best manage and store all this information. It was not until the late 2000’s that the conversations shifted from dealing with the challenges of big bata, to exploring the opportunities for using big data, specifically in commerce.
Defining Big Data
“We define big data simply as statistical analysis of unstructured data”
Before one can fully appreciate and harness the power of big data, one must get past the smoke screen and understand what it really is. We define big data simply as statistical analysis of unstructured data. We see big data less about the data itself, but more about what to do with it. It is less about quantity but more about relevancy – or, how can statistical analyses be used to produce actionable insights, that will in turn translate into business results. A big data initiative does not have to be a daunting, time-wasting, capital-draining pet project. Organisations can build a lean – but effective – big data operating model by involving the right people, instituting a result-oriented process and utilising the appropriate technology.
3 Ingredients to Success
1 – People
The application of statistics by itself is worthless – value is created only when relevant information derived from statistical analyses are used to achieve business results. A successful big data initiative requires not only people who know how to conduct these analyses, but also people who consume, interpret and make decisions based on the results of these analyses. As such, an effective operating model must involve individuals who know how to access the data (information technology skills), understand how to examine the data (statistical skills) and provide context and subject expertise (functional skills). An organisation must empower the right people in order to hold them accountable for reaping the most benefits from its big data investments.
2 – Process
In addition to facilitating collaboration among key talents, a lean big data operating model must also employ a process that both repeatable and adaptable. Before diving into the actual data, an organisation can benefit immensely if it first frames the opportunity and stablishes specific goals. This promotes focused ambition and prevents the organisation from trying to bite off too much. With the boundaries of the initiative defined, the team can then develop hypotheses, identify input data sets and set up analytical models. conduct analytics, validate hypotheses and derive insight. Finally, a successful big data initiative produces not only actions to capture opportunities, but also lessons that will help the organisation refine their operating model, scale up their capabilities and take advantage of big data across its businesses.
3 – Technology
While the topic of big data is broad and generally associated with innovative technologies, organisations must understand that technology is just one of the building blocks to a successful big data operating model. It is never about the latest and greatest piece of software an organisation can afford – instead, leadership need to focus on the most relevant technology it can use for the immediate opportunities it is trying to capture. Big data technologies should be evaluated on how they can address business needs (relevance), solve analytical challenges (effectiveness), and optimize workflow (efficiency).
Conclusion
Organisations looking to capture value from the big data revolution must understand that technology alone cannot guarantee success. When, and only when a big data operating model is supported the right people, a result-oriented process and appropriate technology can stakeholders reap the benefit of this disruptive revolution.