Business Intelligence (BI) has become a strategic tool to help forecast sales
and expenses as well as control costs. BI systems allow to find customer-buying
patterns that assist in understanding them better and avoid sending mails to all
and sundry which is can also earn customer wrath, as they have to go through
un-wanted mails.
Enterprise need to look at various options for a BI solution:
There are various approaches to a BI solution.
- The DWH layer can be completely secured from end user access and all
access can happen only from domain specific marts created from the DWH
repository. - The end user can access any part of the warehouse for analysis and
reporting - The ODS layer can source data into the DWH and also for reporting
The next logical process, after getting clean data into the
warehouse repository, is to build aggregations and marts as the business
requires. This depends on the implementation, where data is already aggregated
in the warehouse repository, thus allowing reporting and analysis directly based
on them. With the BI solution in place for analysis and reporting, the business
user would be empowered with the ability to make prudent decisions based on the
trends presented by the data residing in the warehouse.
The 'single view'
Organizations depend on data. This is an unstated truth, however, its
significance and the impact of its availability and quality is often ignored.
Unfortunately, majority of the organisations have made little effort to
determine the severity of data quality issues and its impact.
Data is a key strategic asset, so ensuring its quality is
imperative. Organizations collect data from various sources: legacy, databases,
external providers, the Web, etc. Due to the massive amount of data variety and
sources, quality is often compromised. There could be many other reasons why the
quality of the data that organisations collect and analyze is so poor. These
could arise from data entry errors, wrong data acquired from an outside source,
and combining good data with outdated records. Integration of systems with
different data standards could also be an impediment. But none is more
compelling than the simple fact that organisations rely on many disparate data
sources to obtain a holistic view of the business.
With isolated technology platforms across various departments
and disparate data stored across several servers and silos, organizations today
deal with these issues on a daily basis. Therefore, it is important for them to
integrate and augment diverse resources within an existing environment, to turn
vast amounts of data from any source and across channels into the usable
knowledge they need to make better decisions.
Importance of quality of data certainly gains significance as
ensuring quality in the data does not only help give a "single and true'
vision of the customer, but also help decision-makers get the high-quality
information needed to determine how best to allocate resources, manage costs,
and more.
It is however important to note that the chosen environment
for data quality must be low-risk, easy to build and manage, and flexible to
change with evolving business needs. The solution should not only automatically
analyze data from a quality perspective, but also give the ability to plan and
justify data cleansing expenditures. By addressing data quality at the source,
data cleansing becomes a proactive rather than a reactive process.
Apart from exponentially improving the quality of the data,
the Data Quality solution will enable organizations to bring together a
360-degree view of their suppliers, organization, customers and enterprise.
With contribution from editorial advisors-Balaji
Krishnan, senior manager, I-Tech Development Team, Oracle India, and Dhyanesh
Joshi, Consultant, SAS India