by Chris Twogood, Teradata
In the world of Big Data, there is an almost infinite variety of information to be explored and leveraged. And, fittingly, there is a huge range of options for Big Data analysis, too. The foundation of Big Data analysis is people examining Big Data sets to seek meaning or patterns to discover business insights or answers to specific questions.
If that sounds pretty open-ended and unstructured, there is a more focused approach to Big Data analysis:
The better, more logical way to conduct big data analysis is by asking business users about the highest value questions no one seems able to answer. Then start looking for the data sets to answer these questions. In other words, start with the questions the business believes to be most critical and then choose relevant data to find those answers.
The point being, Big Data analysis doesn’t have to solve world hunger on the first day. In fact, in many instances, focused analysis on business questions (e.g., what percentage of discount would enable us to move X number of products?) can yield better results in the short-term while growing internal interest and momentum for other long-term, more substantial Big Data analysis initiatives.
Before we dive into different types of Big Data analysis, it is important to stress the critical role of people. Who should do Big Data analysis? There’s no one-size-fits-all answer, of course. But, for most organizations it will be business analysts and managers who understand discrete parts of the business (for example the supply chain or digital marketing). These analytical tools can then be used to dig into data to discover opportunities and insights to improve performance, cut costs or achieve other business objectives. (Note: getting the right people and teams in place is a hugely important but often underrated driver of Big Data success.)
Now let’s talk about some of the different types of and techniques for Big Data analysis. Behavioral analytics is a common – and increasingly essential – capability across a range of industries. As the name suggests, this flavor of Big Data analysis starts with understanding consumers and how they act, why they do what they do, and how their behaviors change over time with the goal of understanding customer behavior can be influenced in the future.
Financial services and retail are two of the sectors that are setting the pace for behavioral analytics. Banks use it to identify suspicious patterns and strengthen anti-fraud capabilities, while retailers are focused on tracking consumer actions across channels so they make the right offer or product recommendation at the right time.
Predictive analytics is another hot topic in Big Data analysis. For example, it’s at the heart of Yahoo! Japan’s effort to tailor advertising and service offerings to customers based on clear insight into their past actions. The results? $100 million in ROI.
Connected or connection analytics is another technique for Big Data analysis that is gaining traction, largely because it’s delivering real value to business users who are engaging it in a variety of very cool ways. Fundamentally, connected analysis is a way to discover patterns across the huge number of connections between people, products and connected devices from the Internet of Things.
The potential applications for connected analytics in particular and Big Data analysis in general are cross-enterprise and cross-industry. HR can hire more effectively by finding useful links between potential new hires and current employees. Sales teams can focus on the right cross-selling and up-selling opportunities and improve marketers’ understanding of influencers. Information security and risk management pros can track cyber security threats more precisely. Maintenance staff can minimize downtime on key equipment and fleets.
There are more types of Big Data analysis we could cover such as data visualization and graph analytics. Each may involve a different toolset or different combinations of data, but the common denominator here is the idea that Big Data analysis converts raw data and information into ideas for tapping into previously unknown opportunities and business insights – for smarter decision-making.