Rapid developments in Business Intelligence have resulted in decision makers and BI practitioners engage with new tools for insightful visualizations. The desired objectives are achieved only after in-depth research and analysis and this makes the use of effective tools important.
There are two prominent approaches to rendering reports and dashboards.
Top Down – Wherein the BI user is aware of the resultant visualisation.
Bottom Up – When the BI user wants to convert the voluminous data into meaningful visualisations.
One must understand that Dashboards and Reports are just the tips of the Iceberg. The multitude of Visualizations and Designs rendered on the surface need predefined, specific and well-architected data structure underneath.
What it takes to create engaging Visualizations?
Data sources are the basic building blocks for any BI initiative. Depending upon the analytical needs, required data sources are identified for their continuous availability and validity. Most of the data needs are fulfilled by internal data sources. Some of the cases where external data is needed are market trends, forecasts, industry benchmarks, etc.
Depending on the need of the data source, there exists the need for data connectors. Most of the time, data is directly available to be accessed from OLTP or intermediate sources which are the exact replica of OLTP real time / near real time data. Some data sources may need proprietary or 3rd party connectors.
One of the most important factors is the location of data. Depending on where the data is available, you may need additional data connectors. Some of the data sources privately available on premise need data gateways to be installed for cloud-based solutions.
ETL / ELT Process
It is obvious that you should not use available OLTP data as it is. The Extract, Transform and Load (ETL) or Extract, Load and Transform (ELT) process helps you clean and transform the data for intended use and as per the transformation needs.
Transformed data is made available via intermediate data houses which store pre-aggregated data for further slicing and dicing operations. Mostly the data that is required is kept in a warehouse and the historical data is archived for later use as and when required.
With increasing need of Business Intelligence, development houses are coming up with powerful visualization tools for enhanced visual experience and seamless data connectivity with various data sources. Now a days, most of the tools offer support like row level security, custom visualizations, data connectors and gateways.
Hosting and availability
Depending on the business need, cloud hosting or on-premise hosting is made available. Given the high availability, flexibility and security, cloud hosting is preferred many times. CAPEX and OPEX considerations also play an important role in this decision.
These are some of the aspects which are of paramount importance in the development of the Dashboards and Reports. Some of the factors which affect decision making are custom visualisation, processing cycle time, accessibility, maintenance and so on.
Someone has rightly said, “What we see on the screen is just the tip of the iceberg”