That is understandable. People just do not like to
share data. Usually, the user department owns both
the system and its data.
The HDB staff had to devise a mechanism of dual
ownership in which the CIO owns the data warehouse
architecture, and the respective user departments
own the data in it. Access to the data warehouse
would be granted to the various users only if
the data owner authorizes it. The authorization list is
reviewed annually, and individual departments are
then at liberty to add or delete users to the list.
Since the data warehousing project started in 1994,
Alex Siow has been actively promoting the spirit of
cooperative learning and doing among the various
departments. Departments are now more willing to
share data, because over time, they have learned that
the data actually belong to the HDB. Their all-important
role is to capture, preserve and store them so
that relevant parties could access and use them for
the common good of the HDB.
5.5. Management issue no. 5 — choosing project
leader
In addition to getting a good combination of
people in the project team, getting the right person to
lead the data warehousing effort is crucial for its
success. The project leader must not only be technically
competent, but should have adequate business
knowledge as well as good interpersonal skills. This
is because people issues which deal with conflict
and politics. are often more difficult to resolve than
technical issues which deal with problems that can
be solved by technical solutions.. In fact, the Head
of the Database Development Unit commented that
‘‘technical issues are within our control. It is getting
different user groups to agree on how different data
should be defined that is difficult. Hence, it is very
important that the project leader has good interpersonal
skills.’’
Previous literature has emphasized the importance
of a project leader who is able to motivate the team
to channel their energies to achieve the project objectives.
This is because a data warehousing project can
be viewed as introducing change in the organization.
As such, some parties may benefit or lose power as
the result of the change. A good project leader would
be able to explain and convince the necessary parties
of the need for change as well as mediate between
various parties if disputes or disagreement arises.
Consequently, interpersonal skills and willingness to
listen to users’ concerns are essential prerequisites of
a good project leader and may determine the cohesion
and cooperation of members of the project team.
5.6. Management issue no. 6 — proÍiding formal,
systematic training
Developing a data warehouse is a difficult endeavor,
but realizing significant benefits is much
more difficult. As such, users must undergo continual,
formal, systematic training to get the most from
the data warehouse. In the case of the HDB, for
instance, the Head of Database Development Unit
said, ‘‘Besides putting in many hours of formal
training, we also follow up with regular ‘hand-holding’
sessions to help users to take full advantage of
the data warehouse.’’ Two factors underlie the need
for user training: 1. the users have a better understanding
of the functions that the data warehouse
will support; and 2. they will be accountable for
making the data warehouse produce timely and accurate
information. Technical system quality is important,
but just as important, if not more so, is the need
to understand the human issues in technical change.
A formal, systematic training program helps to raise
user awareness vis a` vis the possibilities and limitations
of the data warehouse. ‘‘At the HDB, we treat
training costs as investments, not costs. Viewed it
from the investment perspective, management is more
than willing to release training resources,’’ said Alex
Siow.
At the HDB, training is every bit as critical a path
from data warehouse project perspective as having
the warehouse firmly in place. Management realized
this and funded the best training process possible.
5.7. Management issue no. 7 — scalability and data
warehouse maintenance
The HDB operates in a changing environment.
Rules and procedures are rendered obsolete and new
ones implemented. There is no end in the data