John Walker admits there was a problem with trust prior to implementing new data quality software at Bell Mobility’s business customer-acquisition call centre. Employees were not convinced the customer contact data they were using to sign up new clients was completely accurate.
A frequent complaint — one by no means unique to Bell Mobility, rather a common affliction found in the call centre industry as a whole— was that the contact list “sucks,” said Walker, the associate director of customer insights and analysis for Bell Mobility. The result was employees who lacked faith that their contact list would bear fruit.
Walker made the comments at a SAS Institute Inc.-sponsored data quality seminar in Toronto on Wednesday.
Andreas Bitterer, vice-president of enterprise analytics strategies with market researchers Meta Group, agreed. “If the word gets out that the data warehouse does not provide correct answers…then people will not use it,” he said. Bitterer cited the case of a Dutch manufacturer which spent US$150 million on a customer relationship management solution, only to find no one use it for lack of trust in the data.
To solve the problem Walker’s team decided to implement data cleansing software (only to his unit, not across all of Bell Mobility) from DataFlux Corp., a company acquired by SAS four years ago.
Data used in the call centre is gathered from diverse sources such as infoCanada, Bell’s own customer list and Dun and Bradstreet (D&B). The problem is that there is no standardized method for listing customers. “To a computer, a Bill Smith and a William Smith is a different person,” Walker said.
So without clean data a salesperson sometimes ended up calling the same person twice, which not only frustrated sales staff but also alienated potential clients. Duplicate entries were eliminated and addresses were validated using Canada Post listings.
Additionally, with the new solution Bell Mobility was able to virtually eliminate having existing customers on the prospects list. “I say virtually because matching is never 100 per cent,” Walker admitted. Bitterer agreed, adding that that notion, along with the 360-degree view of a customer, is just not possible given the general complexity of systems and data today. Although he did say companies must address, at the very least, the consistency of data on a given customer.
Unfortunately this is not always the case. “Most of the companies we talk to are completely oblivious that they have a (data quality) problem,” Bitterer said. “If some of the CIOs we talk to would take the I (from their title) more seriously” that would solve a lot of the problem, he said. This reality still surprises Bitterer given that information is the “life blood” of organizations.
Back in Germany, where he is based, he often gets frustrated at his own telco, which provides him with wired, mobile, DSL and ISP services, yet often does not understand he is just one customer. “It is actually pretty incompetent on their behalf,” he said.
With Walker’s unit, though, the results with the new system were almost immediate.
“Probably a month after we implemented DataFlux we saw…results jump.” Right party contacts (right phone number for the right person at the right company) went up by 50 per cent, he said. “They (sales staff) started to believe in the list more and they became more motivated,” Walker said.
On the installation side, the process was straightforward, he said.
Often Bell Mobility uses CGI Group Inc. to help install new technology and there is an associated cost to this,Walker added. But for DataFlux, “we had [it] up in a week without any IT involvement,” he said. There was no need to integrate it with any server or database since DataFlux provides direct access to multiple sources, and it just loads onto a PC, Walker said. “It was actually a very easy implementation process.”
Walker’s suggested attendees identify areas within a corporation where a quick win — something like increased sales or reduced customer churn — can be shown to the CEO or CFO to “show the value” of installing data cleansing software.
Bitterer admitted trying to introduce data quality to an entire organization at once is unlikely to work, and that attendees should, as Walker suggested, find the sweet spots and “start somewhere.”
