Kyle McNamara

Creating an analytical capability that allows for rapid model development and analysis requires (among other things) easy access to data with little cleansing. However, in many organizations, customer, product, and transaction data is distributed across systems or databases, making it more difficult to quickly develop a data set for analysis. Last fall, I performed a web analytics project and we were able to use less than half of the data available to us because the data sets came from multiple systems and could not easily be matched. Although we drew some insights from the smaller data set, one of our recommendations was to consolidate the data and make it easier to match so that the next round of analysis could be more complete.

I read today that at RBC, all customer data is owned by the enterprise and held in a central database. I don’t know how Wells Fargo manages its data, but I suspect it is distributed throughout the organization; I have two loans with them, my name is misspelled on one of them, there are different rules that govern how I access and pay each loan – when I interact with them, it’s almost like they are two different companies. If these two banks wanted to target their customers with offers based on product holdings, RBC would quickly be able to understand my customer profile, transaction history, and propensity to buy, whereas Wells Fargo would probably require more time to link data sources to develop the same information. RBC is better able to compete because they already have a central data source for analysis.

Last week, I received a mailer from my gas company pitching air-conditioning options to customers with boilers. (Most homes with boilers lack the ductwork required to install traditional “forced air” systems, but there are other options available.) Granted, my home is over 100 years old and utility companies collect information about services at each premise, but I was still impressed by a utility company’s use of data to develop targeted offers based on what they know about me.

Contrast that with Verizon Wireless. After several years of being a customer, I switched to another carrier – but I wasn’t happy with their service, so I came back to Verizon a few weeks later. Rather than re-activating my old account, they setup a new one, effectively losing the history of our relationship. (When I asked the sales rep if they could reactivate my old account or copy the data to the new one, he responded, “No, we can’t do that,” and proceeded to enter my personal details from my driver’s license.) Verizon can still send me targeted offers, but the extra account history would improve their revenue or churn-prediction models.