IBM BUSINESS ANALYTICS: FIRST TENNESSEE BANK

A detailed ROI case study

Published on 25-Mar-2011

Validated on 03 Jun 2013

Customer:
First Tennessee Bank

Industry:
Banking

Deployment country:
United States

Solution:
BA - Business Analytics, BA - Business Intelligence, Cloud & Service Management, Enabling Business Flexibility, Enterprise Modernization, ROI Study, Information Infrastructure, Infrastructure Simplification, Optimizing IT

Overview

Nucleus Research examined the use of IBM Business Analytics at First Tennessee Bank in order to identify the benefits of applying predictive analytics to marketing campaigns. Analysts found the application enabled the bank to increase profits by better segmenting and targeting its customers, leading to better response rates for the marketing campaigns it uses to cross sell products to customers. Adoption of IBM SPSS predictive analytics also reduced the annual number of campaigns and the amount of time spent on them.

Business need:
the bank decided it needed a better way to analyze the large volumes of customer-related data it was accumulating. Although the bank found it easy to gather data on its customers, it found it harder to use the information to improve decision making.

Solution:
In order to cross sell more of its products to customers, the bank decided to use IBM Business Analytics.

Benefits:
By using IBM Business Analytics to analyze data related to customer activity and historical marketing campaigns, First Tennessee Bank reduced its marketing costs, increased net income, and improved the productivity of marketing staff. ROI: 642% Payback: 2 months Average annual benefit: $899,095

Case Study

THE COMPANY

First Tennessee Bank is a Memphis-based regional bank with 200 business locations in 16 U.S. states, Hong Kong, and Tokyo. The bank provides various financial services to a customer base that includes consumers, businesses, financial institutions, and governments.

THE CHALLENGE

In 2007, the bank decided it needed a better way to analyze the large volumes of customer-related data it was accumulating. The bank had built a data warehouse and gathered data from sources such as online banking records, call centers, and ATMs. Although the bank found it easy to gather data on its customers, it found it harder to use the information to improve decision making. The challenge is one commonly faced by organizations that gather data from multiple customer touch points. If there is no way to analyze customer-specific data and predict customers’ future behavior, valuable data is unused and opportunities to improve decision making will be missed.

One area where the bank saw opportunity for improvement was the marketing department’s mass mailings. Every month, the bank sent out a mass mailing designed to get customers to purchase a bank service they were not already using, such as a checking account or a CD. The campaigns were relatively successful even though they were based on limited data sets and automated statistical tools were not used. However, marketing managers believed their campaigns would have higher success rates if they could target their mailings based on a broader and automated analysis of the bank’s customers: their preferences, behavioral data, and monthly transaction histories.

THE STRATEGY

In order to cross sell more of its products to customers, the bank decided to use IBM Business Analytics, which the bank had already purchased but not deployed. In order to incorporate the application into its decision making processes, the marketing department:
 Trained. Formal training was given to five members of the marketing department, who learned how to build models and perform queries in the application.
 Integrated. The application was integrated with all of the bank’s sources for customer-related data, as well as historical data on the marketing department’s mass mail campaigns.
 Modeled. Using historical data sets, the marketing team experimented with different models in order to both hone their modeling skills and identify ways to improve existing mass mailing practices.
 Consolidated. The banks 12 campaign planning cycles were consolidated to four quarterly cycles after creating a model that enabled staff to make campaign segmentation, materials, and quantities more stable throughout the year.

In early 2007, IBM SPSS predictive analytics was fully adopted by four users in the marketing department and used to design the bank’s quarterly mass mail campaigns.

KEY BENEFIT AREAS

By using IBM Business Analytics to analyze data related to customer activity and historical marketing campaigns, First Tennessee Bank reduced its marketing costs, increased net income, and improved the productivity of marketing staff.

Key benefits of the project include:
 Increased cross selling. The adoption enabled the bank to increase its net income by selling more of its services to existing customers. Before the deployment, the marketing department completed monthly cross-sell mass mailings by segmenting the bank’s customers into tiers based on activity levels and determining which bank products each customer was not already using. After the adoption of IBM SPSS predictive analytics, many new variables related to customers’ preferences and banking habits were included in the design of mass mailings. By analyzing more customer data points, such as their ATM habits, transaction volumes, and call center interactions, the marketing department increased the success rate of these mailings by an average of 3.1 percent, resulting in the sale of more checking accounts, savings accounts, CDs, and home equity loans to its existing customers.
 Reduced campaign costs. Printing, materials, and postage costs for the bank’s mass mail campaigns were reduced by 20 percent. Printing costs were reduced by using the application to decrease the annual number of campaign planning cycles from 12 to four, which enabled the bank to purchase materials less frequently and take advantage of volume discounts. The total number of pieces printed and mailed was also reduced because the application enabled the bank to exclude from campaigns customers who were least likely to respond to a mailing.
 Increased productivity. The bank’s 4-person marketing team reduced the amount of time they spend on marketing campaigns by 8 percent. Two project benefits caused this improvement. First, adoption of IBM Business Analytics made it easier for the staff to perform queries and design marketing campaigns faster. Second, by increasing the effectiveness of their campaigns, the marketing department was able to reduce the number of campaigns they had to build.

KEY COST AREAS

Key cost areas for the deployment included software, consulting, personnel, and training. Two First Tennessee employees dedicated 25 percent of their time to the 4-week deployment and were helped by two consultants who assisted with the configuration of the application and its integration with the bank’s data sources. The bank purchased three concurrent seats and formal 2-day training sessions for the four users in First Tennesee’s marketing department. On an ongoing basis, one member of the marketing department spends approximately five percent of his time supporting the application by performing upgrades, adding new data sources, and assisting users with queries and research. Because of the relatively small footprint of the application, the application was deployed on existing server space and no hardware investments were necessary.

BEST PRACTICES

One reason the adoption was so successful is that early on, the marketing department overcame potential adoption resistance. Prior to the project, members of the marketing department were concerned that changes to the bank’s mass marketing practices might be met with resistance. The problem was the potential reduction in the size of the mailings. Marketing managers knew that when predictive modeling is applied to mass mailings, one result is a higher number of wins, while another is a reduction to the size of the mailings, which might have been viewed as risky. To win stakeholders over, the marketing department used IBM Business Analytics to build projections of the new campaigns, including their costs and expected win rates. Also included was a calculation of the higher ROIs that could be earned on campaign expenditures as a result of their lower costs and higher win rates.

CALCULATING THE ROI

Nucleus calculated the costs of software, consulting, personnel, training, and other investments over a 3-year period to quantify First Tennessee Bank’s investment in IBM Business Analytics.

Direct benefits quantified included profits earned on checking accounts, savings accounts, CDs, and home equity loans sold as a result of the increased success rates of the bank’s mass mailings. For each product line, the benefit was calculated based on the annual number of mailings, the increase to campaign win rates, and the annual net profit contribution from a new sale. The benefit of reduced campaign materials costs was calculated by determining the total cost of printing, paper, and postage for the 12 campaigns completed during 2006 and comparing it to the cost of the four campaigns completed during 2007 after the marketing department adopted IBM SPSS predictive analytics. Indirect benefits calculated included the improved productivity of marketing personnel. This calculation was based on the number of people who save time as a result of IBM Business Analytics, an estimate of the amount of time saved, and their average fully-loaded annual cost. A correction factor was applied to account for the inefficient transfer of time from time saved to time spent on new work.

THE BOTTOM LINE

Nucleus Research examined the use of IBM Business Analytics at First Tennessee Bank in order to identify the benefits of applying predictive analytics to marketing campaigns. Analysts found the application enabled the bank to increase profits by better segmenting and targeting its customers, leading to better response rates for the marketing campaigns it uses to cross sell products to customers. Adoption of IBM SPSS predictive analytics also reduced the annual number of campaigns and the amount of time spent on them.

ROI: 642%
Payback: 2 months
Average annual benefit: $899,095

Products and services used

IBM products and services that were used in this case study.

Software:
SPSS Modeler Server, SPSS Statistics Base

Legal Information

© 2011 Nucleus Research, Inc. Reproduction in whole or part without written permission is prohibited. All calculations are based on Nucleus Research's independent analysis of the expected costs and benefits associated with the solution. NucleusResearch.com

Document options