Mobile operator turns to analytics to beat capacity crunch


Mobile operator 3 has turned to advanced analytics to form a better picture of its network, and to handle congestion issues caused by smart phone users.

In the week that Three joined the mobile operators offering the iPhone 4, the company told IT PRO how it is relying on new data mining tools to ensure its network can keep up.

The system, based on IBM's Infosphere technology, will monitor more than 100 million call data records, as well as details of text and picture messages, roaming, and volumes of mobile broadband data. Each month, 3 estimates that it will process 1TB of compressed data.

The operator decided to update its business analysis tools to tie in with a national marketing campaign that focuses on improvements in 3's network, explained Darren Silvester, information management architect at Hutchinson 3G.

"We are looking very specifically at what subscribers experience on the network," he said. "For voice, that is dropped calls. For mobile broadband and [smartphone] handsets it is the throughput we are getting. Where we see peak loads of data, we can understand where we might have performance issues, and see where we might need to put in more coverage. That feeds directly into the customer experience."

Hutchinson 3G typically looks for a payback on IT projects of under a year, a target which Silvester believes the Infosphere technology will meet "comfortably". The system is also flexible enough, he believes, to be expanded to analyse other parts of 3's business in the future.

At the same time, the analytics tools need to be simple enough for network capacity planners to use, without the need to train to become business intelligence experts too. "The technology does need to be accessible so people can get up to speed on it quickly," Silvester said.

According to Henry Cook, IBM's business intelligence sector manager for Europe, the Middle East and Africa (EMEA), more businesses are turning to analytics, specifically to improve their customers' experience.

"It is about being able to deploy sophisticated techniques, but also making it easy to use and accessible to normal marketeers, so you don't have to have a PhD in maths," he said.