Users’ needs and preference may be expressed in terms
of usage patterns. However, the nature of ISP industry,
where users and management may never see each other face
to face, makes it impossible to develop a traditional indepth
mutual understanding. Currently, the most common
means in servicing customers in ISP industry in Taiwan is
through call center. Over the years, it was found that a call
center is a rather passive way to service customers; it basically
waits for customers to call to present problems. The
most management can do with a call center is to call customers
to understand the reasons for switching to another
company. In addition, management has found the following
problems with a call center. Firstly, staff of a different
shift may not be able to respond properly to the customers
who call at an earlier shift. Secondly, it is difficult to market
effectively, because customers are usually not in a happy
mood when they call. Thirdly, it is difficult to measure
the effectiveness of each individual staff. Overall, most
ISP providers in the nation today are lacking the knowledge
of their customers’ network usage behaviors; they
are not able to raise the profile of customer loyalty. As
Oracle stated, customer loyalty might be the only sustainable
competitive advantage in this very challenging economical
time (Oracle Corporation, 2006). Thus, at this
stage, any company who knows how to deal with customers
effectively will have the definite strategic edge over others.
In addition to customers’ usage patterns, the network
resource management is also an important issue that can benefit both customers and the company. With proper
management of network resources, an ISP company must
plan for resource allocation according to users’ needs of
geographical nature, which will aid in achieving better cost
effectiveness.
The objective of this research is to propose a Business
Intelligence (BI) process for the ISP industry in Taiwan,which could assist management in developing effective service
management strategies. The ISP industry in Taiwan is
characterized by the heavy pressure for raising revenue
after hefty capital investments in the last decade and the
lack of knowledge to develop effective competitive strategies.
The process applies data mining, visualization, and
RFM customer value modeling as the underlying methodologies
to identify various knowledge patterns. These patterns
form the basis for discovering business intelligence,
which includes the identification of VIP status, characteristics
of different usage group, users’ monetary contribution,
and network facility utilization. A BI decision support system
is developed with MVC (Model-View-Controller)
architecture to facilitate the intelligence dissemination.
The performance of the system is empirically and subjectively
evaluated by the company staff. The remaining structure
of this paper is as follows. In Section 2, we present a
review of underlying methodologies that are utilized in
the study. Section 3 describes various phases of the proposed
BI process. In Section 4, we implement BI process
with actual data from the company and describe the development
of a BI decision support system. The system evaluation
is described in Section 5, and Section 6 concludes
this paper.