Managing Information Technology

• Running the data center out of an expanded
  Indianapolis facility would save an estimated annual
  $2 million or more in occupancy costs due to the sig-
  nificantly lower costs in this Midwest city.


• The new Sallie Mae data-center facility in Reston
  could be leased out at an attractive price.
  Shortly after the relocation decision, a lease agree-
  ment with a new tenant, beginning July 1, 2001, was
  signed for the Reston data center.

The First Data Center Consolidation

In the months prior to the USA Group merger, Sallie Mae
had been working to integrate the operations of Nellie
Mae, a company Sallie Mae had purchased in 1999. Nellie
Mae was a major originator of student loans based in
Braintree, Massachusetts, with 150 employees. However,
the integration project had been slow going. The 16 IT
professionals at Nellie Mae already had severance pack-
ages in hand, but their severance pay was dependent on a
successful operational move and knowledge transfer to
Sallie Mae.
When Jo Lee Hayes of USA Group took on this small
data center consolidation project, severance of IT profes-
sionals was only 3 months away. Working side by side with
Nellie Mae’s IT operations head and a team of six to eight
people, Hayes defined the current Nellie Mae systems,
established a move strategy, and worked through the sys-
tems integration and knowledge transfer issues. Successful
on-time completion of this data center project was a visible
early win for the Indianapolis-based IT team.

Everybody knew where he or she stood, and there

was no question about what needed to be done. We

had a hard-and-fast date, and people were motivated

to combine operations successfully because sever-

ance was tied to successful knowledge transfer....

It was thrilling, impossible, and such a rush when we

actually pulled it off.

—Jo Lee Hayes, Vice President, Business

Solutions Group

Critical Application Decisions

The types of factors used to determine the fate of current
applications included system functionality, scalability, per-
formance, the number and types of interactions with other
systems, whether the system was custom or purchased, and
if purchased, whether the latest version of the application
was currently installed.

Now you’re talking about jobs. Now you’re talking

about changing people’s lives in a big way with

whatever system is selected.

—Allan Horn, Vice President, Technology

Operations

The analyses under the two assigned champions (one

for Sallie Mae, one for USA Group) led to one of three

outcomes:

• A consensus recommendation was reached amicably.


• A consensus recommendation was reached, but after
  much conflict and strife.


• A consensus recommendation could not be reached,
  which meant escalating the decision to the executive
  level.

Champions were technology owners who had a

chance to show top management that they could make

decisions and execute them. It didn’t reflect well if

the decision had to be turned over to senior officers.

—Jo Lee Hayes, Vice President, Business

Solutions Group

The choice between the two custom-developed loan-

servicing applications was the most contentious decision

and had major IT work force effects. Larry Morgan, head of

application development for USA Group at the time of the

merger, had been responsible for managing the develop-

ment of both loan-servicing systems. Morgan first joined

Sallie Mae in the mid-1980s: he was hired away from

Pennsylvania Higher Education, where he had helped to

develop a loan-servicing system being used by both Sallie

Mae and USA Group. Due to Sallie Mae’s fast growth, a

new loan-servicing system (Class) had to be developed and

installed within a 2-year time frame. In 1991, USA Group

lured him to Indianapolis to build a system with even more

functionality (Unity), installed in 1994.

At the time of the merger, Sallie Mae’s Class sys-

tem was a 15-year-old application written in COBOL

and CICS, maintained by more than 75 IT people in

Reston. The only major change to the system had been

converting it from an IMS (networked) database to

IBM’s DB2 (relational) database. USA Group’s Unity

system used a network IDMS database supported by

Computer Associates.

The gap analysis that was performed for the two

loan-servicing applications (see criteria in Exhibit 2)

did not result in a clear choice. Although there were some

concerns about the scalability of the IDMS database

application, the functionality of the Unity system was more

Case Study IV-2 • FastTrack IT Integration for the Sallie Mae Merger 615