Big data splits Telstra in two

Telstra has formed a cross-functional team to identify use cases for big data analytics that are both commercially attractive and acceptable to privacy-conscious customers.

The carrier is taking a "fairly softly softly" approach to big data and is in the process of working through possible use cases, chief architect of information and corporate systems, Mark Kortink, told CEBIT's DataCon conference in Sydney last week.

"We've got a lot of people in Telstra with an interest in big data together and half of the people in the room are privacy people — which gives you a feel for how seriously Telstra takes privacy," Kortink said.

"The reality is the privacy conversation almost dominates the big data conversation within Telstra. There's a lot of use cases you could dream up that we just won't go near, won't touch.

"Without getting into a lot of detail, there have been quite a lot of scenarios where a use case has been looked at and then rejected purely on the basis that it's just too 'creepy', to use a word that's being used.

"Telstra values its brand and the trust in its brand way beyond any opportunity big data can deliver."

Kortink saw potential opportunities for big data as being more internally-focused, such as around fraud detection and network optimisation. The latter could include predicting when certain parts of the network are likely to fail.

He said the "general feeling" within Telstra's executive was of interest in the potential of big data, but that there still needed to be a good business case presented in order to go down that path.

"What we've tended to find is we're looking for use cases all the time, but no one has really put up the serious use case where people have said 'it's got a good business case attached' or it's worth putting the effort into developing a really good business case and taking it further," Kortink said.

"In effect, we're still looking for the list of opportunities we're going to pursue."

Kortink said that big data infrastructure would be treated separately to Telstra's substantial existing investment in data warehousing.

"[Data warehousing] is basically a reporting platform — it's not an analytics platform," he said. "In the target architecture we're looking at, we're treating analytics and big data capabilities as new capabilities that we need to add into our reporting and analytics environment."

The potential for big data was, on paper, huge.

"Telstra has probably got more data than anyone else," Kortink said.

"We've got mountains of data and we've got areas where we do quite a lot of fairly smart analytics at the moment and always have.

"I think big data just gives Telstra a whole lot of opportunities to improve the analytics it already does and take on analytics it couldn't do in the past because it was too expensive or too slow."