Page 510 - Handbook of Modern Telecommunications

Network Organization and Governance 4-41

on time and proactively. The ability to “close the loop” by sending proactive information and forcing the
telecommunications company to act is a challenge that is mired in technological complexity.
Cause: Reporting solutions lack comprehensive alerting capabilities and are not flexible enough to
allow users to define alerts themselves. Even if these facilities are available in some reporting solu-
tions, business users require technical assistance to extract the information they require. Applications
do not provide access to common business definitions and calculations that would allow business users
to define seamless alerts and have the information delivered on multiple touch points.

4.3.5.5 It Takes Too Long to Build a BI Solution
Despite considerable spending on a BI infrastructure, most telecommunications companies find that
what they have does not match up to their unique business needs. Building a BI solution for a telecom-
munications company requires extensive understanding of the telecommunications domain and how
business users perform various analyses.
Cause: The effort to build a BI solution frequently begins from scratch, with teams trying to under-
stand business needs and building suitable models. This can consume significant resources and time.
Teams building a BI solution lack domain expertise. Often, teams of modelers and architects spend too
much time understanding the business. Most vendors offer tools and technologies for point solutions.
There are few complete, connected, and consistent analytic solutions.

4.3.6 Overcoming Barriers
Building a BI solution is more than an implementation of a bunch of technology tools. It requires a
framework driven by a set of guiding principles. Some of these core principles of implementation are
explained here:
1. Establish the need for trustworthiness: A BI solution has to deliver data that is trustworthy. Data
becomes trustworthy when it can be substantiated. Ensure that data that is delivered to busi-
ness users can be substantiated with a seamless traceability all the way to its source, without any
additional effort. When a business user looks at a number and wants to authenticate it, the task
should be as simple as a click of a button. Users should not have to depend on technical resources
to draw out simple reports. In short, the traceability and lineage of data should be comprehensive
and seamless in a BI solution.
2. Establish the need for a holistic view: Build a BI solution to provide a holistic view. Even though
the effort may begin with a single department or division or a specific subject area, the solution
should be designed such that additional subject areas can be added seamlessly. For example, a
telecommunications company can begin by building the BI solution for risk. However, the tele-
communications company must ensure that the underlying risk structures are created such that
profitability, CRM, and the investment sides of the business can be accommodated at a later date.
If efforts are decentralized and managed by individual departments, it is important to ensure
that they coordinate the underlying structural aspects in ways that can facilitate extensibility and
integration. Unless such strategic guidance is available, efforts tend to be independent.
3. Establish the need and sanctity of consistent business terms: A BI solution without this fundamental
driver is set to fail. Facilitate and create a glossary of terms that is consistently applied across your
enterprise. It is important to involve business users in this exercise. Also appreciate the diversity
among departments, and ensure that everybody’s requirements are accurately described.
4. Enforce consistency at all costs: Defining or establishing a business language in itself does not solve
the problem. A mechanism to enforce this common language must be established. The ability to
enforce this feature across the BI solution is a key component of the solution. Telecommunications
companies must realize that departments would prefer managing their own subsets of data and
would like to limit their access to those subsets. Departmentalized subsets are inevitable and are
simple to manage and maintain. The challenge is to control their proliferation.