Come the “big data revolution”, business intelligence (BI) analysts will be out on their ears – or so some people have argued.
But as Computing’s recent IT Leaders’ Forum on “Assessing the Future of BI in Big Data” heard, the reality will be very much more complex.
The key difference between big data and BI isn’t just scale and the use of structured and unstructured data to glean insights, but also that it offers real-time or near-real-time analytics that can be acted on by the business. BI, in contrast, is typically focused on analysing historic business data.
Hence, the two disciplines don’t just require very different IT skills, but very different attitudes and approaches, with BI staff today typically able to offer a deep understanding of the business, according to Sandeep Sachdeva, vice president in the global business information practice at Sogeti.
“In the Netherlands, police use telco data to predict where a crowd is going to be, so that they can make sure they have the resources to control the crowd, if necessary,” Sachdeva says.
In contrast, big data enables an organisation to analyse data in a more outward-looking, customer-centric manner – very often in real time or near-real-time.
In retail big data makes possible just-in-time advertising and just-in-time couponing to attract customers as they walk into – or past – a store.
Media agency MEC Global, says CIO Trevor Attridge, can take weather data to McDonald’s, telling it when and where it’s going to rain, which means that the fast-food restaurant chain will sell more coffee – rain means that people will be seeking both shelter and a nice, hot drink.
Free Wi-Fi initiatives help tap into this, by enabling an organisation to identify customers that use their Wi-Fi, which typically request such details as email address and mobile phone numbers, enabling offers to be sent straight to the customer.
As a result, therefore, says Andrew Maclaren, a global big data consultant with Brilliant Data, BI and big data require very different skills. But this does not mean that BI is redundant. Far from it, the people working on BI, he says, have the kind of skills that organisations will need in the future.
Furthermore, says Lauren Sager Weinstein, head of analytics at Transport for London (TfL), there remains an entrenched distrust of computer-crunched data in many quarters. TfL, for example, has traditionally used surveys to find out about people’s changing routes, but even today when it would be quicker and more accurate to interrogate the detailed fares data generated by the Oyster Card e-payment system, there is resistance.
“‘It’s not as reliable as our survey data that we have that’s five years old’,” Weinstein is told by the traditionalists at TfL. “But millions of taps coming in [via the Oyster Card] is better information than you’d get from a survey,” she says.