Dan Matthews discovers what it takes to understand the masses of data now available – without getting in a muddle
“Data, data, data; I can’t make bricks without clay.” True enough, but this is not the exasperated cry of a 21st-century, Bluetooth-enabled executive, but that of Sherlock Holmes, enduring an agonising period of down-time in The Adventure of the Copper Beeches, first published in Strand Magazine in 1892.
Data, big or small, is nothing new. It has been curated, considered and manipulated for millennia and companies have long used it to make all kinds of decisions. The faddish rise of big data is not down to an increase in the overall amount of information, more that datasets have become more accessible and, with the right tools, ordered.
Its growth is inextricably linked to the internet, which lets people deposit information about themselves into a range of digital troves, some secret and some open for all to see. Without really thinking about it, anyone using the web leaves a breadcrumb trail of information which gives clues about the things we like and how we act.
For companies this is great news: understanding the habits of consumers enables businesses to fine-tune their products and marketing materials, making them leaner, more efficient, more relevant and, best of all, more profitable.
The stumbling block for many businesses is that big data feels inaccessible and over-complicated, like quantum mechanics or the film Mission Impossible. It is considered something only massive, moneyed organisations, such as governments and blue-chip behemoths, could bend to their will.
To the rest of us, it’s time-consuming gibberish – a numbery nonsense. Yet while mega-businesses are collectively meeting the big data challenge, firms outside the FTSE 250 have a golden opportunity to out-smart their competitors by investing in better information.
The first step on the path to this happy place is to get over the name. “Big data” combines two words that, while intimidating separately, conjoin to make us all want to run and hide. Simon Shorthose, managing director of ReadSoft UK, says it’s better to think less “big” and more “lots of small”.
“Start by finding all the pools of small data – they’ll be there,” he says. “Then consider how to get all those smaller pools working together. Having masses of data is fine, but what are you really looking for from your data?
A big data-enabled organisation should give thought to investing in some seriously high-calibre geekage
“When there is a reliable source of information and a strong thought process behind that information, people in your company will come to rely on it. But if it’s just a mess of data, then people will not change the way they make decisions.”
Now that you have a sense of perspective, there is a six-point plan to follow, according to Paul Alexander, chief executive and co-founder of Beyond Analysis. First, be clear on your business objectives and those of your customers. Second, understand what data you have right now (and what you don’t). Third, align points one and two, and then create your own big data strategy. Fourth, plan in relatively simple detail for the first 12 months, then build more complexity in years two and three. Fifth, only include actions that can be tested and measured so you can learn and develop. And sixth, be committed to your data – if you’re in it for the long haul so will your customers be.
“Big data is simply about joining the dots of all your relevant data sources,” says Mr Alexander. “Whether it is customer, local-market or communications data, every company out there has at least one data source at its fingertips.
“The mistake many companies make is trying to bring all of it together, all at once. What they should be doing is taking the time to understand what their version of big data should look like for their particular business, pick one area of focus, and commit energy and resources into delivering against it.”
Big data is not just for consumer-facing firms either. Examples abound of businesses honing in on datasets to improve what they do. Take Rolls-Royce, for example, which captures performance data from flights and uses it to anticipate when parts will need servicing or replacing.
Another, albeit smaller, business making good use of the data available to it is GRITIT, a major winter-gritting company working across Britain. It sends nightly jobs out to gritters via smartphone, tracking the work done and creating reports in real time.
Brendon Petsch, IT director at GRITIT, says: “We have a huge amount of data, drawn every night during the winter, and there are literally millions of variables. We provide our service out of hours and it isn’t always obvious we have been on site. This data on file is, therefore, vital to us.”
But if anything is out of place, the devil will be in the detail. “By analysing the data, we can see if any variables are wrong and the teams can investigate if necessary, even on a night when thousands of site visits took place,” he adds.
Generally speaking, the “data opportunity” is causing increased demand for systems architects and analysts capable of capturing and processing the data into chunks everybody can understand. A big data-enabled organisation, therefore, should give thought to investing in some seriously high-calibre geekage.
“Big data requires a different mix of abilities than exists in most organisations today,” says James Riley, global head of innovation at HCL Enterprise Application Services. “A greater number of information analysts will be needed to convert the information into intelligence. Through an appropriate training, these skills can be developed internally.”
Those who get it right, he says, could ultimately create organisations that think more like people than inanimate objects. “As individuals, even when doing the simplest tasks, we constantly process a wide variety of information. Organisations are not nearly so sophisticated and certainly nowhere near as close to real time,” says Mr Riley.
“Big data, and the step-change in processing capabilities it provides, has the power to make organisations act more like humans who constantly seek to manage trade-offs and optimise, rather than wait for the end of the month to figure out what went wrong.”