BMO Bank of Montreal Works with SAS® to Foster Knowledge Creation

Strategy and knowledge management are inseparable at BMO Bank of Montreal, according to Carl Touchie, managing director of decision support services. SAS software fits into that maxim by enabling advanced automated systems for using customer data to its most profitable advantage.

BMO Bank of Montreal is Canada's oldest bank and functioned as the country's central bank up to 1935. It is one of the largest in North America, with over US $160 billion in assets and 32,000 employees.

In 1996 the card-management department embarked on an integrated, fundamental shift in strategy from portfolio profitability and management to customer profitability and management. The bank developed micro segments, based on such things as purchase behavior and the "share of wallet" of individuals. This micro segmentation added complexity and gave rise to a need for process automation and integration of the scoring processes, hence the bank looked to SAS for the technology to accomplish this, coining it the Automated Scoring Application (ASAP). This new strategy certainly seems to have been successful: BMO Bank of Montreal's Cardholder Line of Business' net economic profit has increased 89 percent over the past three years.

Results achieved by the implementation of the SAS application so far include time-saving analytical capabilities, automated and much faster score creation (from days to hours), a move from manual output review to exception reporting, and an unlimited number of supported scores per modeller. Using SAS for ASAP was a natural step, as SAS has been in use at the bank for more than 20 years.

A Fundamental Shift
BMO Bank of Montreal's drivers for a comprehensive knowledge management strategy are familiar to anyone involved in banking: The competition was starting to heat up. "The U.S. banks, such as Capitol One, MBNA and Banc One were investigating expansion into Canada and we wanted to be sure that we had a defensive strategy in place," says Touchie. "We needed to understand customer behaviors and change our focus to delivery of actionable customer knowledge. Our Knowledge Delivery Process (KDP) is defined as the process of providing and receiving information in the form of a score or a set of scores to either a process or technology that has the ability to influence or customize the customer contact channel to create value."

The bank set up a knowledge creation approach looking at customer data through four different perspectives:

• Financial drivers, of which the primary application is overall portfolio performance management.
• Purchasing behavior, which refers mainly to enhancement and merchandising development.
• Share of wallet, which is concerned mostly with retention and balance transfer.
• Seasonal, which refers to time frames such as Christmas and the back-to-school season, and was intended for the timing of contact strategy.

There are three types of scores in the system. Classification or segmentation scores assign customers into groups based on a set of common characteristics, such as financial behavior. Evaluation scores assign customers a type of financial value, encompassing historical profitability and lifetime expected value. Predictive scores are an attempt to present the likelihood of behavior such as campaign response or churn.

BMO Bank of Montreal's automated scoring application became a major component of the KDP. SAS is the core technology used by the bank to support data manipulation, model development, automation of data flows, and to maintain version control of scoring routines.

"Our objectives in creating the automated scoring application were in the revenue, performance efficiency, and risk and control areas. We wanted first of all to improve channel effectiveness, thereby increasing revenue. Automating our score generation process now happens via shareable and reusable data objects and we have improved resource utilization via parallel job scheduling, and so have reduced manual intervention and risk of error. It was also important that we segregate duties between score development and implementation and produce audit reports," says Touchie. "This application really works for us because our extraction logic is written once and stored centrally. This facilitates the parallel extraction of data. SAS also takes care of the complex data extraction process, which is a benefit to the modeller. We now have ease of maintenance and transfer of knowledge of all our models."

Touchie explains the application's extraction logic. "What would happen before we had this application was, we would be looking across a number of models and extracting certain data fields for them individually. Now what we do is schedule the different data fields that we need to extract for different models and the extractions can be reused, which significantly reduces our extraction time. Another big advantage is that the application is automated and will perform extractions and run algorithms according to a set schedule, which requires no manual intervention. We don't have to worry about things like the analyst not being there on time."

When BMO Bank of Montreal runs a profitability analysis, the application retrieves all key customer data that is attached to the algorithm required to calculate the profitability. The algorithm is then within its process and takes the data, fills it in, calculates, and gives the profit score. Then the algorithm calculates the real profit, and all the profit of all the customers is added up in order to give a figure for the profit of the aggregate of customers. "A business rule established by the bank will say that we expect the profit to be XX million every month, and within 5 percent of XX million one way or another. If it is not, the rule stops the process and we'll go in and examine the situation. In the past, we would have had to go in and actually look at all of this information and try to figure it out manually," Touchie explains.

Time savings is quite a good benefit in BMO Bank of Montreal's new application: running the account segmentation model has gone from 22 hours to two, the profitability model from 24 to four, and the soft attrition model has gone from 24 to three hours.

The automated scoring application has 20 power users, including statisticians, modellers and analysts.

Working with SAS
"We tend to go to SAS when we have a specific kind of issue and work together in partnership to find a solution," says Touchie. "To give an example, the automated scoring project is really a boutique solution that came out of us saying, 'Here's problem that we have.' We had a few ideas of how to solve it, and SAS had some basic data warehousing processes that could form the basis for the solution. So, I think that's essentially being innovative with them."

A significant factor in the success of this partnership is the understanding that exists between SAS and the bank. "Our SAS account managers tend to be very good at getting into our organization and networking within the business and understanding some of the problems we have. They come up with solutions for us that come from their understanding of our business. I feel like we have a mutually beneficial relationship with SAS," says Touchie. "SAS tools integrate very well, so once you have the SAS knowledge, you can create applications that exactly suit your requirements."

BMO Bank of Montreal's technology investments are calculated to help the bank support its business focus. "We had to change the whole thinking process in the credit card organization. Even once you've made a change, some employees will still revert to talking about products rather than being focused on customer behaviors. We are simply interested in customers – not even customer segments, necessarily – individual customers. We are focused on giving customers a card that meets their needs," says Touchie. "I think we're on the leading edge of innovative use of technology, the knowledge management process. The automated scoring process is a very important tool in helping us to make a basic cultural change at the bank – from products to customer focus. Our SAS technology helps us make our business strategy a reality."

Copyright © SAS Institute Inc. All Rights Reserved.