fields, how do you do the change management, etc.”
Chile, however, decided to curtail the training to the
markets embarked in this initiative. They said, “it is
now your responsibility to get training”... [I]t was
a nightmare. We had to rely on one particular
resource that had been an integral part of the imple-
mentation in Chile, Luis Boesch. He became very
important. Definitely, he was an ombudsman.
Initially we were so reliant on him that he could not
leave even to provide consulting to Colombia, which
was the other market implementing MaxFli.Through a long and difficult process of examining
screens, reports, and documentation, the MaxFli project
team began to form ideas about the inputs, outputs, and
processes involved in MaxFli. However, apart from inter-
actions with Luis Boesch and participation in several elec-
tronic discussion groups about MaxFli implementation,
there was no way to verify their conclusions.
The BATCA project team struggled to find qualified
IT personnel to support MaxFli in each country. While
stronger economies like Panama and Costa Rica had ade-
quate IT personnel, it was more difficult finding qualified
IT personnel in struggling economies like Nicaragua and
El Salvador. Consequently, the MaxFli project team hired
and coordinated all development staff in Costa Rica. To
confirm the capabilities of applicants, Gracia had appli-
cants complete a small programming scenario along with
their written application for employment. Those applicants
who successfully completed the programming module
were further considered for hiring. This process allowed
Gracia to identify qualified personnel whom he believed
added value to the development process. Ultimately Gracia
was very satisfied with the quality of his team.
A final challenge faced by the project team in BATCA
was turnover on the BATCA executive steering committee.
Six months into the MaxFli project, the GM of BATCA
transferred to Europe, and a new GM, Raymond Acorda,
was selected from Souza Cruz, BAT’s Brazil operation.
Then shortly after the project rollout, a new marketing exec-
utive, Juan Morales, joined the steering committee. While
both Acorda and Morales supported MaxFli, the mid-project
leadership change created additional difficulties for the
implementation team.
Implementation in Central America
The MaxFli implementation in Central America struggled.
It was completed on time but over budget. Much of the
budget over age was caused by the necessary transition
from MaxFli 1.0 to MaxFli 1.1. This transition was
required because MaxFli 1.0 was designed to handle a
single currency and a single management structure. It
could not support the complexity of multiple countries and
multiple currencies required by BATCA’s cluster structure.
The software development challenges associated with a
major system upgrade were made more difficult by anoth-
er transition occurring with BAT globally.
In May 1999, Globe House decided to phase out all
projects with Andersen Consulting. To continue the develop-
ment and support of MaxFli, BAT began a joint venture with
Cap Gemini/Ernst & Young, called Ciberion. In September
1999 formal responsibilities for the ongoing support, train-
ing, and development of MaxFli were given to Ciberion.
Ciberion was created to develop, market, and sell the
MaxFli system to BAT end markets and external companies
in the consumer goods industry. The transition was difficult
for Chiletabacos and BAT Colombia, but especially difficult
in Central America. To Gracia, Ciberion seemed too focused
on future sales to pay attention to the present needs of BAT
MaxFli users. Although Ciberion had committed to a 2-week
turnaround for MaxFli problems, even “urgent” support
requests took an average of 60 days to resolve. Gracia
recalled waiting for weeks to hear from Ciberion even when
his staff had already developed a solution to the problem.
Project IT Leader Gracia felt that several require-
ments that BATCA considered important were neglected
by Ciberion. At least three are worth mentioning. First, the
disagreement about the importance of MaxFli training for
the implementation team caused continual frustration.
Despite continual requests for training from Kruger and
Gracia, Ciberion insisted that it was not required. This had
long-standing implications for the success of the MaxFli
implementation in Central America. As clearly stated by
Rodrigo Palacios of Chile, “MaxFli is a business change
program supported by an IT program.” Without proper
training, the BATCA implementation team was at a severe
disadvantage in implementing the system.
Second, by the time BATCA implemented MaxFli it
was clear that there were problems with the Jornada hand-
helds used by MaxFli. Both Chiletabacos and BAT Colombia
had experienced problems with the limited durability and
unreliability of the Jornadas. The BATCA project team felt
that these difficulties would only be worse in Central
America. Central America has a tropical climate with regu-
lar, often daily, periods of heavy rain. The project team felt
that the climate might be too stressful for the handheld.
Additionally, many of the sales routes in Central America are
in remote, rugged, and isolated locations. Gracia was con-
vinced early on that the handhelds were not adequate:We assessed the handheld ruggedness and therefore
knew that it was not robust enough and was going to
bring lots of trouble to our sales force. However, sinceCase Study IV-6 • The Challenges of Local System Design for Multinationals 657