BT, the UK’s main telecoms operator, is developing and using advanced business intelligence systems to help keep customers as it faces increasing competition in the market and tackles regulatory pressures.
The firm is continually setting up and employing advanced systems, largely based on off-the-shelf business intelligence technology from Oracle, to tackle individual aspects of its business that it would like to improve.
Most notably, in the past two years BT has made major developments in how it uses business intelligence to improve the effectiveness of its call centres in retaining and ‘upselling’ to customers.
Simon Griffiths, chief designer business intelligence and data warehousing at BT, told delegates at Butler Group’s Business Intelligence Symposium in London that since 2006 BT had completed three major projects in that field, enabling its 15,000 call centre agents to quickly access key data and avoid the pitfalls of searching through the company’s 15 terabyte database of customer information.
The first, called Next Best Activity, had the goal of giving call centre agents clear guidance on what to offer customers phoning in with queries or complaints. Since September of last year, BT has provided agents with a simple screen showing customer behaviour and suggesting sensible offers to upsell to customers or reasonable discounts to offer in the case of complaints.
“We have a difficult balancing act of selling customers what they want and making sure we sell what we need to ensure strong profit,” said Griffiths. “This system helps our agents make those decisions and we’ve seen over a 10% increase in sales since implementing it.”
Customer Viewer, the second project, was implemented at the same time and had the aim of reducing the number of customers BT was losing to rivals in the highly competitive consumer broadband market. The system provides agents with key customer facts and behavioural data in one system, avoiding the need for agents to have to run multiple windows on screen.
Griffiths explained that the technology was used to help agents decide how much time and how many steps should be taken to try to keep each customer that had phoned in asking to end their service: “They use the screen to work out, ‘Are you worth keeping and what can we offer that’s attractive to you?’”
It was particularly important for BT to quickly decide which customers were key, he explained, because regulatory bodies had made it reduce its market share. “As we had to lose customers, it was important we lost customers we weren’t making a lot of money off, not the high value ones,” he said.
The third project, implemented this year, was called Merlin, and aimed to carefully monitor call centre effectiveness across teams and down to an individual level, enabling supervisors to guide agents on how to improve their selling.
The system carefully analyses calls, orders, sales and timesheets to provide managers with a clear overview of performance each day and take remedial measures. It has also enabled BT to replace extensive legacy systems with what it called a more agile setup, including Oracle 10g databases, Oracle Business Intelligence Enterprise Edition, Hyperion Business Intelligence, and HP Integrity Servers.
BT had learnt that it was not necessary to have real time reporting in all situations, but rather to offer “right time” reporting, which was more cost effective and made sure the data was received and analysed according to the timescales of each department, Griffiths said. This could mean 15 to 30 minutes later in many cases, or even up to a month later for financial departments that did not require instant analysis.
The firm had also learnt to ask a wide range of users what they thought of the systems, instead of “IT going off into a corner and coming up with their own systems”, he said.
BT’s systems were designed by a team of over 10 in house engineers, and the production was offshore. Average pre-installation testing times were up to a week, with pilot roll-outs and surveys of users before full implementation.
The success of these projects has led high level management, including non-IT executives, to support the use of business intelligence to tackle more areas including setting new call tariffs.
“We’d like to tackle our core pricing next, based on advanced intelligence of customer behaviour that we have carefully analysed,” Griffiths said.