Continuous validation, understanding the customer relationship, linking behavioural and attitudinal data, and leveraging operational data are the keys to successful database marketing. These are the core learnings gleaned from Wells Fargo’s database of 40 million customers by its senior vice president Gary Class, who addressed the ADMA Global Forum yesterday.

“Even with lots of big data tools, it’s still very hard to understand all the dimensions to the data,” Class said, calling the challenge of vast amounts of data a “curse of dimensionality”.

Presenting at day one of the Forum, Class shared the American bank’s secrets for successful database marketing.

The first tip was to leverage data already in a company’s possession to understand the nature of the customer relationship. “Model the customer relationship to create a sense of who you’re dealing with,” Class advised.

“In banking the customer isn’t always a person. Often the customer is a married couple, others in domestic relationships, or small business and a group like the boy scouts. That complicates the data so we spend a lot of time modelling to understand the relationship with people.”

Building on this point, Class told delegates to transform operational data into relational data. “Use information about payments that were made, data on customer interactions, all that data that created originally to keep track of things. We’ve re-purposed that data to learn about customers.

“Turn it into relational data, so rather than general information about how a website is performing or how the mobile app is being used, you have information on which customer downloaded the mobile app and how many times they used it over a period of time. Repurpose it and orient it around the customer, if you can over a very long time series, to give you rich data to understand customers.”

The bank also focuses on linking behavioural data with attitudinal data, to link what people do with what they say. “We have a research department, a strong culture of customer satisfaction tracking, email surveys and so on,” Class revealed.

“We had a lot of results we didn’t quite understand until we were able to graft it to the customer behavioural data. How long they’ve been with the company and the products they have has a huge influence on their attitudes.”

Finally continuous improvement and data validation was identified as core to Wells Fargo’s success. “There is a non stop need to constantly validate data,” Class said.

“Data can often be messy or corrupt. You have to go to the source of the data; it’s really important to talk to the people that spawned the original data, because often the documentation available in the data warehouse is wrong and has nothing to do with what’s being populated. Really focus on data validation and know how the data was created.”