This year, humans will create or replicate around three zettabytes of data. To put that into context, if you were to convert that data into text and print it as books, the resulting stack of books would stretch from the Earth to Pluto and back more than ten times.
For businesses, that means big opportunities – and big headaches. The super-abundance of data makes it possible to spot new customer trends, work faster, make smarter management decisions and boost innovation. But only if your employees know how to find the nuggets of gold under the mountain.
Most companies have too few analytics-savvy workers. According to Corporate Executive Board, only 38 per cent of employees – and 50 per cent of senior managers – have the ability to make good decisions based on data. As Hal Varian, chief economist for Google, famously pointed out: “The sexy job in the next ten years will be statisticians.”
Against this backdrop, we’re seeing the emergence of the “data scientist” – someone who has mathematics, statistics, machine-learning or coding on their CV, plus the harder-to-teach characteristics of curiosity and a willingness to explore huge amounts of data. “They ask the kind of questions that no one else has thought of,” says Chris Hillman, principal data scientist at Teradata.
Data scientists have to turn abstract data into images people can relate to and, importantly, act upon
So how can you turn your employees into data scientists – and turn your data into profits?
Speak ‘business’ not ‘tech’
“Most people think data scientists should be tucked away in a dark corner of the IT department and report to the CIO [chief information officer]. They’re wrong,” says Harvey Lewis, research director at consultancy Deloitte. “The role of a data scientist is to solve business problems, not technical problems.” The trick is to find people with programming skills and on-the-ground business experience. They need to be able to analyse the “bits and bytes” of data, and then interpret their results to a non-technical audience. Data scientists should report in to a chief data officer (CDO). Fundamentally, the CDO is tasked with being the “voice” of data and representing data as a strategic business asset at the executive table.
Create a data culture
If you’re staffing up now to meet the challenge of big data, chances are you’ll end up stealing talent from your competitors. But getting data-minded employees through the door is only half the battle. You need to send your data scientists on regular training programmes to keep their skills topped up. “There’s no such thing as a data boot camp but there are plenty of courses out there,” says Teradata’s Mr Hillman, who has an MSc in business intelligence from the University of Dundee and is studying part-time for a PhD in data science. Most software vendors, such as Tableau and QlikView, offer classes. IBM launched its Big Data University in October and since then more than 18,000 students have enrolled in its online courses. “Once you’ve developed the data ‘black belts’ within your firm, make sure they filter their knowledge down,” says Girish Pancha, executive vice president and chief product officer at Informatica. Run regular forums where you showcase big data best practice and get people excited about the results. Tiffany, the world’s second-largest luxury jewellery retailer, holds year-round workshops on analytical techniques, for example.
Data scientists are essentially story-tellers or artists. They have to sift through hordes of incoming information to discover a previously hidden insight and then communicate to the rest of the company how that might provide a competitive advantage or address a pressing business problem. “Half of all employees find that information from corporate sources is in an unusable format, so data scientists must learn to present data in a way that quickly encapsulates meaning,” says Deloitte’s Mr Lewis. “They have to turn abstract data into images people can really relate to and, importantly, act upon.” This is where the creativity comes in. In big data speak, it’s known as visualisation and there are plenty of tools on the market – from Karmasphere, TIBCO and SAS, for example – to help. Think animations, interactive graphics and heat maps. Remember, it’s not how big the data is, it’s what your employees do with it that counts.
Don’t operate in silos
Imagine you’ve just taken out a mortgage account with your bank. Two months later, that same bank sends you brochures about its new mortgage products. It’s annoying, right? Many companies hoard mounds of data in departmental silos or business units. Having multiple – and often inconsistent – versions of the same information lurking around the firm can result in embarrassing mistakes, so make sure your employees work from a “master” data management system
Share data with your suppliers
Opening up your data to your suppliers – often known as “teaming” – means your staff can respond to problems as they happen and adjust processes accordingly. “If you’re a manufacturer turning out faulty products, you can use big data analytics to work out what’s going wrong,” says Mr Lewis of Deloitte. “Are all the failed products coming from the same supplier or the same machine? How was that company or machinist briefed?” You can use predictive analytics to better forecast demand and improve stock control, and that means big savings. The Centre for Economics and Business Research chalks up a potential £46 billion in efficiency gains for UK industry between 2012 and 2017.
Retailers can use big data to track a customer’s behaviour and note what life stage they’re at; whether they are married or divorced, a student or have children? Businesses can work out customers’ propensity to buy certain products. They can then use that information to recognise when the customer is nearing a purchase decision and nudge the sale by bundling preferred products, and offering rewards or discounts. “It can be tempting to exploit data in a way that benefits your company,” says Mr Lewis. “But just because you can, it doesn’t mean you should.” With great power comes great responsibility. Data scientists need to keep the customer in mind. They need to be transparent about what data is being used and how it’s being used. “Don’t dehumanise the data,” he adds.