Transform Data into Insights
I used to work in a local telco, where we built applications to support the business and pricing analysts in revenue reporting, product pricing, etc.
In order to support the in-house analysis applications built, we relied extensively on the data which come from the corporate data warehouse. Each month, we built data cubes (incrementally) for the analysts to perform slice and dice operations. One of the applications which the corporate-side wanted us to build was a business scorecard or dashboard kindof web-based system which allows them to monitor the health of the business and monitor key-performance indicators. While we wanted to make use of the s/w used for building the cube to support the web application, we realized we couldn't. One of the main reasons was that unless we buy the "BI portal" from the same software vendor who supplied us with the cube/clients, we cannot interop with the cube directly. Due to cost-constraints, we had to build the application from scratch, replying on complex aggregate queries/etc to show the KPI, as well as support other reporting functions. The time to build these application could range from 3-4 months.
During the annual workplan sessions, one of the directions given by the directors to all the senior managers was to find ways to "create value" from the data residing in the data warehouse. The question is "how"? The next question is when these data are then made available, how can IT provide end-users with tools to visualize, to analyse and more importantly to understand the figures presented?
Transactional data are created each day through the call switches, retail outlets, etc and these data are then pumped into the data warehouse on a periodic basis. The transaction data is rich in details! For example, it could show at a glance the state of the business; trends; interesting customer demographic information; churning patterns, etc etc.. To the IT department, these are just data, that needs to be managed. How can we "create value"? IT department would then need to work with end-users (analysts who might know what they are looking for) in order to create new killer-applications that can move the business into the next stage.
The "truth"/"patterns" etc lies in the data residing in the data warehouse. However, the problem is the very large amount of data available. How can the analyst know what to look for, except perhaps comparing revenue (across quarters, across product groups, across account managers, etc). Experienced analysts could then ask IT to build applications specific to support them in order to establish better pricing, identify key trends, etc.
Herein lies the value of the new BI s/w that are increasingly being made available to end users. Using these tools, IT could eventually "create value" from the data they managed. Using these tools, end-users are now better equipped to seek the "insights" that they need in order to anticipate market trends, and truly transform data into insights.
In order to support the in-house analysis applications built, we relied extensively on the data which come from the corporate data warehouse. Each month, we built data cubes (incrementally) for the analysts to perform slice and dice operations. One of the applications which the corporate-side wanted us to build was a business scorecard or dashboard kindof web-based system which allows them to monitor the health of the business and monitor key-performance indicators. While we wanted to make use of the s/w used for building the cube to support the web application, we realized we couldn't. One of the main reasons was that unless we buy the "BI portal" from the same software vendor who supplied us with the cube/clients, we cannot interop with the cube directly. Due to cost-constraints, we had to build the application from scratch, replying on complex aggregate queries/etc to show the KPI, as well as support other reporting functions. The time to build these application could range from 3-4 months.
During the annual workplan sessions, one of the directions given by the directors to all the senior managers was to find ways to "create value" from the data residing in the data warehouse. The question is "how"? The next question is when these data are then made available, how can IT provide end-users with tools to visualize, to analyse and more importantly to understand the figures presented?
Transactional data are created each day through the call switches, retail outlets, etc and these data are then pumped into the data warehouse on a periodic basis. The transaction data is rich in details! For example, it could show at a glance the state of the business; trends; interesting customer demographic information; churning patterns, etc etc.. To the IT department, these are just data, that needs to be managed. How can we "create value"? IT department would then need to work with end-users (analysts who might know what they are looking for) in order to create new killer-applications that can move the business into the next stage.
The "truth"/"patterns" etc lies in the data residing in the data warehouse. However, the problem is the very large amount of data available. How can the analyst know what to look for, except perhaps comparing revenue (across quarters, across product groups, across account managers, etc). Experienced analysts could then ask IT to build applications specific to support them in order to establish better pricing, identify key trends, etc.
Herein lies the value of the new BI s/w that are increasingly being made available to end users. Using these tools, IT could eventually "create value" from the data they managed. Using these tools, end-users are now better equipped to seek the "insights" that they need in order to anticipate market trends, and truly transform data into insights.
posted by SQL Server Gems at
11:52 PM
0 Comments:
Post a Comment
<< Home