Blending Data to Target Banking Consumers With Pinpoint Precision

I remember the first day I learned what a “JOIN” statement was in a programming class in college. I remember thinking this is going to be really useful someday. If you are not familiar with the SQL JOIN statement, it is a programming function allowing you to combine two or more tables of data in a database to act as one, typically with a common key. For instance, if you could join a table listing the pet preferences (i.e. dog, cat, fish) of individual customers with a table showing which customers do or do not have a credit card you could use the data to drive personalization of a YouTube campaign promoting dancing pets and credit card offers. Nowadays I wouldn’t describe the ability to join data sources as “useful,” I describe it as critical. Below are two examples of website projects in which we used combined data sources to drive website improvements.

Example #1: Joining survey data with Core Platform customer attributes.

Over the years I have been curious about the redundant nomenclature that banks and credit unions use to describe their products. For instance, the phrases mortgage, home loan, and home financing can all be used to describe virtually the same products. Extractable ran a survey with a large F.I. in the midwest to learn which phrase was preferred as well as compare the data with trends on major search engines (i.e. Google) and internal site search usage. What we found was that the phrase “mortgage” is used 7X more frequently than the other terms. We decided to join the survey data with some demographic data in the F.I.’s core platform. We joined the two data sources via the email address of the survey takers. The findings were interesting:

  • First time home buyers 30 years old and younger are actually more likely to call it a “home loan.”
  • Home buyers older than 30, or any home buyer that has purchased a home before, are much more likely to call it a “mortgage.”
  • Customers that are looking for a second concurrent mortgage, (i.e. a rental property), are much more likely to use the phrase “home financing.”

This simple joining of two disparate data sources, (a customer survey and the core platform), yielded some great insights to drive personalization on the website, mailers, email campaigns and informed display advertisements. Taking one step further and combining the phrases with the F.I.s conversion rates on mortgage products with different age groups lets us know how valuable each phrase is in pay per click traffic acquisition.

Read More: Banks’ Power to Mine Data Frightens, Intrigues Consumers

Example #2: Joining Web Analytics with Customer Relationship Management data.

Working with one of the largest retirement plan administrators in the country we spent a good amount of time diving into the web analytics from their secure portal where clients log in to perform retirement planning activities, (i.e. enroll, change distributions, view balances and performance, set up retirement goals, etc.). We had a great opportunity to join the web analytics behavioral data with their CRM data joining the two data sets by the visitors internal account identifier. With this combined dataset we could start to see interesting data such as how online behaviors differed between visitors that are close to retirement or very far away from retirement, (by age or savings). Also, we were able to see how content/tool usage changed between people with investments that were performing well (i.e. 20%+) or people with poor performing investments (i.e. -10%). Below is a list of some of the insights we found:

  • The closer a client is to retiring (by age), the more frequently they visit the site. Inversely, the farther a client is from retiring (by age), the more engaged they are with retirement planning content and tools. From this we could consider developing retirement planning content for younger audiences.
  • Clients >65 years old are 2X more likely to use a tablet than a smartphone. Combine this data with the previous point and we know which types of articles can be written for specific devices.
  • Clients that login to the portal via smartphones and clients that login to the portal via laptops are essentially separate crowds. Less than 3% of the clients use both devices. Although the number of clients that login via multiple devices in growing rapidly. Website owners usually assume that all audiences have similar content and functional needs, but are data shows they are distinct and likely have unique needs.
  • The wealthier a client is, the more they frequent the site – until they become millionaires and then frequency of visits drops significantly. Inversely when they become millionaires, they visit less frequently but consume much more retirement planning and investment content than before.
  • The duration of visits and the pages viewed per session decrease as clients investments perform better. It appears that when times are good, the clients feel less of a need to check balances and read investment content.

These examples provide great insights on how to drive personalization by age group, by investment performance, and by affluence. Combining these data sources to create more focused client segments will allow us to optimize the site tools, content, and traffic acquisition campaigns more effectively.

Each time we combine reporting from different tracking systems we discover new patterns that can be used to inform the development of superior user experiences. It is very exciting to think that as technology platforms become more interoperable in the years ahead it is likely that all forward thinking organizations will become experts in customer preferences and behaviors which will lead to better web experiences and much better service from innovative financial institutions.

This article was originally published on . All content © 2024 by The Financial Brand and may not be reproduced by any means without permission.