Skip to main content

7 steps to gain superior big data insights

The term 'big data' has been likened to a lot of things, one of which being a rubbish dump. And if big data is basically the immense accumulation of digital refuse, left over from billions of daily transactions and interactions, that's okay. There are treasures to be found in that trash.

Analytics experts are quick to point out that the big deal with big data platforms, like Hadoop, is not about size. It's about the value of the insights that can be gained from both large and small data, provided you know how to find them. To that end, here are seven steps you can take right now that will yield better big data insights.

Strategise an end game

Many executives and marketers look at big data analytics as a magic bullet that can cut through mountains of information, target hidden insights and hit them dead-on, setting off lights and bells of brilliant understanding. The truth is that insightful analytics isn't about finding random things that might have some relevance; it's about understanding what you hope to learn and then strategising, being sure to select the analytics tools and techniques that will optimise your chances of finding what you're looking for.

Contextualise the data

As the saying goes, you can't get something out of nothing. And when it comes to getting insights from analytics, this is especially true. We all know that it's difficult to determine the real meaning of words taken out of context. And this also holds true for data. Contextualising data by adding in related data makes better insights possible in a number of ways. For one thing, adding context makes it easier to spot highly correlated data that can lead to insights, as opposed to disparate data that is less useful. Context also makes it easier to make sense of anomalies. Additionally, patterns within contextualised data are easier to recognise than those hiding within non-contextualised data landscapes.

Visualise the information

Data visualisation brings data to life, which in turn brings better insights. Dry analytical results presented visually allow us to easily see what is relevant among all of the variables. Through data visualisation, patterns are captured and communicated plainly and simply, and trends can emerge that may have otherwise gone unnoticed. Visualisations make results easy to share and quickly bring teams together on the same page, thus allowing collaborative insights and breakthroughs to more readily occur. A number of vendors, including Tableau Software, have great enterprise-grade visualisation software.

Articulate insights

Once an insight has emerged, it's critical to be able to articulate it by boiling it down to one simple statement. It's not unlike the movie pitch a producer uses to convey the big bankable concept of their film to the studio execs in as few words as possible. Insights, reduced to tiny post-it notes and prominently placed, can play a big role in helping executives and marketers to visualise the 'big picture', as it emerges.

Defend or dismiss detected patterns

While detecting patterns in data can be a key to gaining insights, it's important to recognise that not all patterns are relevant. Therefore, a careful review and discussion of each pattern must be conducted to see if the potential implications and possible importance of the pattern in question are worthy of defending or if they should be dismissed as irrelevant. Thoroughly scrutinising data is a crucial step in transforming information into actionable insights.

Take a break from the 'board'

Once insights have been recognised, articulated, and thoroughly discussed, an important next move is to step away from the 'board' for a day or two and let them incubate. Upon returning, you'll be able to take a fresh look at things, a look that will either confirm what you already knew or lead to changes that can make the insight even more relevant and valuable.

Stick with what works

Big data and Hadoop can be intimidating, prompting some to feel the need to learn new analytics techniques in order to deal with large data sets. This assumption couldn't be farther from the truth. In the end, big data is still data, and it can be analysed efficiently and effectively by sticking with those analytical tools and techniques that have already been proven to be reliable.

Whether your organisation is sitting on petabytes, terabytes or gigabytes of data, there are valuable insights to be gained from data analysis. By using the right tools and techniques, along with these steps, you will uncover new and better insights to help your firm maintain a competitive edge for years to come.

Michele Nemschoff is vice president of corporate marketing at big data platform solutions firm MapR Technologies.