Blogs

Firms ‘must respect privacy’ when using predictive analytics

The increasing use of predictive analytics tools will be something that many businesses look to implement this year, but one expert has warned they must think carefully about how to best use this technology and keep the practice under control.

Research vice-president at Gartner Frank Buytendijk told an audience at the Gartner Business Intelligence Summit in Sydney that organizations should not only think about implementing tools to help with their decision-making – which may actually be the easy part – but what the consequences of these solutions could be.

"It's easy to use tools [but that] will not lead to better decisions. If you don't know what you are doing, easy to use tools lead to more spectacular failure," he said in a speech reported by CIO.com.

Mr Buytendijk cited research from Gartner that estimates that around a quarter of companies using consumer data will face a reputational challenge by 2016 because they have failed to fully understand the privacy and trust implications that comes with this technology.

Businesses that take their use of data too far by "crossing the creepy line" could see their big data deployments fail and their relationships with customers suffer as a result.

One high-profile example of this could be LG, which was recently forced to apologize after it emerged that the firm's smart TVs were sending data on their user's viewing habits back to the company, even if individuals had opted out of the programme.

Mr Buytendijk also warned that firms need to be wary when using predictive analytics that they are able to interpret the results they get carefully and not allow their own biases to cloud their decision-making.

"We humans are notoriously bad when understanding probability. When asked more than 80 percent of people feel that they belonged to the best 50 percent of car drivers," he said.

Therefore, recognizing which patterns are useful and when they are not is one of the most invaluable skills firms can have when using big data – but getting past the built-in biases that all humans have could be one of the biggest challenges, Mr Buytendijk continued.