Business Intel
Definition: Business Intelligence (BI) is an umbrella term applied
to methods, applications and technologies used to gather, integrate, present and
analyse business information to improve decision making.
| |
|
|
Business Intelligence Solutions (BIS) lead to better business decision making
through providing access to
enterprise data for easy analysis against Key Performance Indicators (KPIs).
This is achieved through having more information available at all levels of an
enterprise and enabling each management level to be more responsive to current
market trends. Every aspect of the business can be co-ordinated efficiently and
dealt with at various levels of management.
As technology has improved, the volume of information available for analysis has
increased significantly and more efficient systems have been designed to handle
the
data collection process. The
data collection services and tools available ensure even microscopic pieces
of information are included for analysis whereas they would have been ignored
previously due to not being cost-effective to collect.
BIS plays a strong role for an enterprise of any size. The development of
automated collection tools has helped reduce the time cost and monetary cost of
intelligence gathering. A smart business will look at evaluating every piece of
data individually and collectively to help make more informed decisions. BIS
enables collective data to be analysed for trends and then for every subset of
data to be drilled down and analysed individually.
Business Intelligence comprises the following main elements:
1. Analytics
2.
Customer Relationship Management (CRM)
3. Dashboards
4. Data Warehouse
5. Data Integration
6. Data Management
7.
Data Mining
8. Extract, Transform and Load (ETL)
9. Online Analytical Processing (OLAP)
10.
Business Performance Management (BPM)
11. Reporting
12. Scorecard
Implementing BIS in Enterprise
There will be a significant cost of implementing any BIS if existing
applications exist for any part of the overall process. The ideal scenario would
be to use solutions for each aspect of BIS from the same vendor or where they
are proven to be able to integrate with other vendors. In some cases, it may be
more cost-effective migrating existing processes into a bespoke system to
facilitate better control and understanding. This will reduce the
training costs
associated with training new personnel as they will only be required to learn
one system as opposed to multiple existing systems.
After the decision has been taken to implement business intelligence solutions,
a set of criteria must be addressed from the outset in order to gain the most
benefit. The most important areas for consideration are:
1. Response Time
Considerations will need to account for data capture time, ETL processing time,
caching and reporting time in addition to user expectations. For example, if the
service is offered to clients and they are informed statistics are updated in
"real-time", they do not expect to be waiting for a few minutes every time they
login to check stats.
It would be unwise to have every piece of data update in "real-time" as it may
cause too much load on the server and result in reliability issues. Instead,
only the absolute necessary pieces of information should be updated in
"real-time". This area needs to be clearly defined in the design process - what
constitutes essential information.
2. Data Refresh Rates
Automated queries and database dumps can be routinely scheduled to take
snapshots of actual statistics to reduce potential problems with the live data
capture. Providing only the vital information is extracted, it can be set to
retrieve a database dump every minute if necessary. The data integration tools
should be designed to have minimal impact on a server when importing new data
sets to facilitate smooth data exchanges. If designed correctly, management
would be querying
backup data from the latest data refresh rather than causing additional load
on the live server.
3. Visual vs. Analytical Dashboards
Statistical information is essential to represent the current state of affairs
and should be available for drilldown analysis. However, a visual
dashboard should suffice to present a quick overview of affairs with any
changes highlighted. It is especially useful for CEOs to see actual
performance against KPIs instantly.
4. Data Delivery
It isn't necessary for every managerial level to have access to all the data
collected, but it is necessary for them to have access to all data relevant to
their decision making. In this instance, BIS must be designed from the outset to
have flexibility in assigning different roles. A bespoke solution that enables
new roles to be created where specific sets of data can be extracted and
delivered for analysis without needing to cross reference with other departments
is essential.
5. Scalability
After an initial assessment of enterprise requirements it is still important to
consider scalability issues and possible future requirements. Any BIS
implementation should adequately provide the capability for future modification
and expansion without posing any significant risk to current procedures and
management requirements.
The criteria above is not extensive but does cover the most common
considerations that can be overlooked when designing bespoke BIS applications.
The more thorough the planning undertaken before designing a bespoke solution,
the more useful and cost-effective the end solution will be.
The future of Business Intelligence
With the advent of new technology enabling more efficient data capture and
processing, traditional business intelligence has started to shift from reactive
to proactive in the sense artificial intelligence (AI) can aid decision making
when certain conditions exist. This helps to free up more time for managers to
focus on other areas like staff motivation and training.
There will continue to be a need for managers to analyse data against KPIs, but
the majority of decisions can be automated through use of AI and alerts sent out
for manual intervention if the data produces any anomalies. The role of
artificial intelligence should be defined from the outset in order to be able to
manually disable or adjust it according to enterprise goals. Failure to
adequately define the AI role can lead to a loss of control in the decision
making process.
Overall, business intelligence and BIS form an integral part of every enterprise
and if used correctly will help improve efficiency and help meet both short-term
and long-term objectives. (See
open source solutions to Business Intelligence)
|