News Stay informed about the latest enterprise technology news and product updates.

Integrating network data with BI systems broadens customer view

Adding network data to existing BI and customer data analytics systems can help telecoms, ISPs and others improve customer service and make networks more efficient, according to a recent report.

Recent estimates predict that YouTube will bring in around $240 million in revenue through 2009. Unfortunately for Google, which bought the popular video-sharing site in 2006, it will spend $700 million in operating expenses during that same period.

Much of YouTube's operating expenses go toward maintaining the site's mammoth network connection that allows it to store and distribute millions of user-created videos. The site, like many Internet service providers (ISPs) and telecom carriers, must balance network performance and customer service with revenue and profitability.

To do that more efficiently, telecoms and ISPs should consider mining their network data for key customer information, like dropped call rates and download activity, and combining it with the more traditional business intelligence (BI) systems to allow for more in-depth customer analysis , according to a recent Yankee Group report.

Jon Paisner, an analyst with the Boston-based research firm and author of the report, said that the result should be a more complete view of the customer and the network, giving carriers the ability to better allocate network resources, respond to customers' network-related complaints, and up-sell customers on new services and products.

"The idea is to do data mining of current network activity to create actionable information for the operator," Paisner said in an interview. "It gives you the ability to plan, predict, optimize and reconcile [customer buying patterns]."

Most telecoms and Internet carriers currently collect and analyze customer data to identify their most profitable customers and those customers most likely to leave for a competitor, also called customer churn. But the data being collected, which includes billing information and number of calls to customer service, for example, often doesn't include network activity, Paisner said.

That can lead to an inaccurate picture of a customer.

More on business intelligence software

Read about PivotLink's new business intelligence Google gadget  

Learn how one school district overcame a business intelligence deployment gone horribly wrong  

Find out whether business intelligence vendors and customers are seeing eye-to-eye

If a customer downloaded three ringtones and made no complaints to customer service last month, a wireless carrier might deem that customer satisfied. But maybe that same customer experienced a number of failed ringtone downloads before succeeding, owing to poor network performance. The customer is probably not satisfied at all and may be more likely to switch carriers if the difficulties continue, Paisner said. A sales rep analyzing traditional customer data via BI software or tools would have no way to know this.

Integrating network data with customer data for BI can also help carriers identify likely periods of increased network activity so they can plan to add excess capacity. It can also help forecast the impact that a new, large customer would have on network performance.

The problem, however, is that network data is often collected by different systems even within the same company, with one system used to manage network provisioning and another for customer activation. Breaking down those silos, then consolidating the network data within a data warehouse with other customer data for BI and analytics, is an extremely expensive undertaking, Paisner said -- one most companies are unlikely to undertake in the midst of a recession.

"The idea of complete convergence of all network and business data into a single repository with universal analysis capabilities is an unattainable dream," Paisner wrote in the report. "The siloed nature of data today, combined with the distinct nature of network and IT data, means that operators will always have to mine multiple data sources for true end-to-end intelligence."

What telecom and Internet carriers can do, he said, is make targeted efforts to consolidate network and customer data around specific service or application types – voice data, video data, managed services -- through data federation techniques. There are even some vendor partnerships, like the one between Agilent and Teradata, to help carriers develop such targeted network and customer data consolidation efforts, Paisner said.

When deciding which network data to focus on, "operators should examine how existing databases interact with key processes and existing infrastructure today," he wrote. "They should pick the data locations that require the highest overhead to move, and consolidate around them. Not only will this maximize operational efficiencies, but it will lessen the initial pain and disruption that data convergence can cause."

Once targeted network and customer data consolidation is achieved, carriers should also create a dedicated policy management layer where carriers can apply the insights network data provides, including the ability to make changes to both network and subscriber policies, Paisner said.

"This functionality can lie in a variety of network elements," he wrote, "but it must be present in some fashion to enable operators to make the most of their newfound intelligence."

Dig Deeper on Business intelligence best practices

Start the conversation

Send me notifications when other members comment.

Please create a username to comment.