The term “big data” is now common in the business world, where it is seen as a potential game-changer. Much is said about how it could help businesses reduce costs, increase efficiency and create a competitive edge, yet scratch the surface and people may not actually understand what the term means. Jim Banks speaks to Jardine Lloyd Thompson Group’s Ian Cohen about how companies should use big data to best effect.
It is often the way that important concepts in the business world become marketing buzzwords before they are fully understood. Such is the case with big data, which has rapidly become a familiar term, but is a concept to which business and academia are only now waking up to.
Ian Cohen, CIO of Jardine Lloyd Thompson Group (JLT), is not afraid to express provocative views on new technology trends, and he is outspoken on the subject of big data, which he sees as holding immense value that very few have been able to unleash.
"It is a hugely misunderstood term. It is in the marketing mix with concepts like cloud computing and other buzzwords. That is OK, as it is the way of the world that buzzwords precede the enabling use of any technology," he remarks.
"It is unfortunate that marketing terminology leaps ahead because big data is a hugely important topic. It demands more considered thought because it is hard to define. In fact, I would say it is worth trying not to define it, and worth not being distracted by the marketing language, because there is and always has been data, whether it is small, big, deep or shallow," he adds.
Cohen's point is that business should not necessarily look at big data as a new concept, but should build on its existing understanding of how to extract value from data.
"We have more ability to amass, aggregate and interpret data than ever, but there has always been data, even if there are new algorithms we can apply to it. You may not be making the best use of it and you may not know where it all is, but it has always fallen into the same categories," he explains.
"There are the things that you know, the things that you would like to know and the things that you don't know but may stumble across serendipitously. A lot of focus has been put on the last of these categories, with people using Google searches to spot flu outbreaks, for instance. But that is just 20% of the value of data, and the other 80% comes from what you already know, but are not using effectively."
Ask the right questions
Albert Einstein famously said that information is not knowledge, and the challenge with big data is to derive knowledge from a seemingly limitless amount of information. Knowledge, after all, is power.
For Cohen, one way to derive valuable knowledge from big data is to avoid being caught in the trap of looking ever further into the data field, as the risk is that one will become lost.
"Use what you already have, but use it better. Deriving value is about interpretation rather than analysis, so interpret and make the interpretation visual if you can. I'm keen on visualisation as a way to spot patterns in a mass of data. That is how you see new opportunities," he says.
"With big data, it is easy to get hung up on a buzzword, or to be seduced down a path of limited value. You need to understand what you are looking for, and be able to analyse data and interpret it quickly to give it meaning. Among all this data, the most important skill to have is the ability to ask the right questions. Data analysts are still important, but the valuable new skill is interpretation. You need to know how much data you need to make the decisions that drive outcomes."
Cohen's point is that because a business can collect data about everything, everywhere, it must decide whether doing so is really valuable. A business must also understand the responsibilities that go along with the type of data it collects, and should also have clear intentions in terms of what it wants to achieve by interpreting that data.
"You need boundaries. A lot of emphasis is put on finding serendipitous value in data, but if you pursue clear goals by asking the right questions for your business then those serendipitous insights will still come along the way," Cohen remarks.
"There are huge upsides to big data. For instance, if in-car telematics can lower my kids' insurance premiums because they are driving safely, or if they can predict patterns in driving that help to modify behaviour, then that is valuable. That kind of outcome-based data collection has a lot of value," he says.
Getting big data wrong, however, could see corporates wasting money on unnecessary investments in technology and analysis. To avoid this, they should know exactly what they want from their data.
"To be something different, you must first define what it is you want to be. Then you can alter behaviour in a way that moves the organisation towards that new state. Don't try to change before you know what your goal is. First, know what you want to be, even if the definition is loose to begin with," Cohen urges.
Back to basics
One way in which big data is expected to help businesses is by providing better insight into customer behaviour. Many organisations talk about the value of microsegmentation of their customer base as a way to improve their product or service offerings. Cohen believes this may have value, if done in the right way, but that there may be easier ways of addressing this issue.
"I see many retailers wasting time and money looking for patterns in the behaviour of their customers. In fact, customers can talk about their experiences more than ever, so those retailers could just start out by listening to what customers are saying," he says.
Big data is not a myth, nor is it a siren luring companies to their doom with an illusory promise of great rewards. But neither is it the solution to all problems. It is a concept with great potential, but to deliver on that potential, companies must build on their existing data strategies, and extrapolate the skills and competencies they already have.
"There is no pixie dust to sprinkle on the problem. It is the same legwork as before, just with better tools. Remember, it is not magic, it is just data. Define what you want to achieve and be rigorous about pursuing it," says Cohen.
"Companies already have 95% of the data they need, so they should focus on outcomes such as delighting their clients or gaining market share. There are some huge and positive upsides from interpreting data effectively. The serendipitous upsides that everyone is looking for will come on a journey towards a tangible outcome."