Bank of Montreal
Bank
of Montreal develops "MIND" for enhanced business
While it had a wealth of data
about its customers, their accounts, investments and service needs, the
Bank of Montreal did not have an easy way of providing this information
to its management team. What was needed was a decision support system
that would supply information tailored to all of its decision-makers,
ranging from top executives to line-of-business (LOB) managers to branch
managers.
APPLICATION
PROFILE |
Project: Management
Information — New Direction (MIND)
Purpose: To
give Web-based access to business performance information such as
customer information, product profitability and revenue reports.
Benefits: Data
can be mined from the bank's customer database, analyzed and presented
to diverse groups of users in customizable reports delivered over
the Web.
Platform: IBM
RS/6000 SP servers
|
To meet this need, Jan Mrazek,
senior manager, business intelligence solutions, and a team of IT professionals
developed Management Information — New Direction (MIND), a Web-based information
super mart running on top of the bank's Customer Knowledge DataBase. CKDB
is a 2T enterprise-wide data warehouse that consolidates highly detailed
information on all of the bank's customers. This includes the month- ly
profitability of more than 18 million accounts, individual customer profiles,
product categories and their performance, and profitability of different
market segments and regions.
The enterprise-wide customer information
data warehouse runs on an IBM RS/6000 with the DB2 Universal Database
Enterprise-Extended Edition for AIX. MIND utilizes the MicroStrategy Intelligence
Server, Web and Agent for its database query and reporting tools.
The MIND project was developed
to run on top of the CKDB — mining, analyzing and presenting data to users
in customizable formats. Users are now able to get much more detailed
and extensive information on product profitability and customers than
ever before. Information can also be customized to give the most value
to each user. Users include financial service managers, LOB managers,
senior executives and human resources. Each group has a different front-end
interface to accommodate presentation style and user preferences. There
are currently 500 users using MIND, with a total of 2,000 expected by
the end of 2000.
TOOLS |
IBM DB2 Universal Database Enterprise-Extended
Edition for AIX; MicroStrategy Intelligence Server, Web and Agent;
Platinum ERwin; Platinum ModelMart; AB Initio
|
Kathy Lisson, executive vice president
and chief internal sponsor of the MIND project, convinced all of the division
vice presidents of the importance of taking an enterprise view of management
information systems rather than splitting them up along individual lines
of business. Reporting to both the CTO and CFO, she understood the value
of having quantitative data and analysis for business decision-making.
The success of the CKDB data warehousing
project, delivered in July 1998, went a long way in convincing top management
that the MIND project was feasible. The CKDB offers its own data mining
and analytical reporting capabilities, which enabled the bank to achieve
a 100% ROI on the project in just six months.
The MIND project began in October
1998 with a three-month pilot in which Mrazek and his team had to prove
the feasibility of their plan before additional funding would be granted.
DEVELOPMENT
TEAM |
Patricia DaFre
Vanitha Devarajan
Jan Mrazek
Ted Mendes
Doug Welch
Linda Lee
Stella Chan
Sanjiv Shah
Ken Ma
Roy Cheng
Eric Chan
Victor Halfkenny
Andre Tardif
Nelio Lucas
Steve Hoskin
|
One of the challenges the team
had to deal with was the project's large data volume and user base. Several
models were built to see how users would access the system and how MIND
would perform under presumed loads. "We refined the models as more information
was acquired during the development process," Mrazek said.
To safeguard the bank's data,
MIND has six levels of security. Users' rights to information are set
when they log on with their personal ID. The system shows users only the
information they have rights to.
Users can analyze demographics
to view product success by market segment or region. They can also look
at customer demographics, or see a total picture of a single customer's
association with the bank.
This enterprise-wide picture of
the bank's customers has created opportunities for cross selling and specialized
marketing. In addition, looking at product performance at an enterprise
level has led to product revisions and enhancements.
Although customers do not have
access to MIND, it provides the bank with a significant CRM tool by giving
bank executives and managers better information. For example, Mrazek said,
an investment manager using MIND might notice that a customer has $500,000
in an account that is earning very little interest; the manager can use
the available data to contact the investor and help place the money in
a safe but better performing account.
Another significant CRM benefit
of MIND is the ability to identify customer problems early on. For example,
the system might notice a customer moving some of their many accounts
to another bank. A manager could contact the customer to see what could
be done to retain the accounts. Human resources can also be better allocated
based on trending done with the system. Busy customer times can be spotted
and extra tellers added to those shifts.
The first phase of
MIND was delivered in June 1999. Debugging and system stabilization continued
until September 1999. In phase two, Bank of Montreal hopes to pay more
attention to feasibility studies and to provide additional contingency
time between major project tasks to allow for unforeseen circumstances.
— Rich Seeley
KEANE
REPORT
|
Each of the entries in the
Data Warehousing category demonstrated innovation in aspects of
the project; however, the Bank of Montreal demonstrated a level
of creativity that was unmatched by the others.
Two factors that contributed
to the success of the pro- ject were ensuring end-user involvement,
and an internal sponsor that was involved and committed. But where
Bank of Montreal's creativity gleamed was in the actual development
of the project. Examples of innovative use of resources include:
- Leveraging RAD techniques
and Web-based information delivery mechanisms to overcome the
challenge of geographically dispersed end users.
- Developing their own
software development methodology based on best practices and consulting
engagements.
- Testing performance and
scalability by modeling the final system based on anticipated
usage.
- Leveraging AB Initio
to enhance data transformation.
The development accomplished
all of this while managing an effort that encompassed 40-person-years
of effort. All of the judges remarked that this project, and the
techniques used, clearly demonstrated innovation and discipline
in achieving a goal.
Team Leader:
Nat Bongiovanni, senior principal consultant with Keane' e-Solutions
Architecture Practice
Team Members:
Luise Kroeger, senior data warehouse architect with the e-Solutions
Architecture Practice; Nageswaran Puliyadi, senior principal consultant
with Keane' e-Solutions Architecture Practice; and Subhash Rastogi,
senior principal consultant with Keane' e-Solutions Architecture
Practice
|