Key Activities of Operational BI
In this post we will cover the key activities of operational BI and define its key characteristics.
Operational BI helps managers to focus and optimize critical business processes daily. The primary purpose is to access appropriate data in nearly real-time fashion (these days it is even possible to start a discussion about real-time delivery) to be able to make decisions that can influence a positive business outcome.
Normal use-case might include users looking at the operational data through a dashboard or a client copying data from the DB and importing it to Excel for further analysis.
The table below examines the most important characteristics of Operational BI:
Immediately, one notices that the focus is on the right-now data to drive real-time decision-making. There is also a focus on time sensitivity – data being pulled from the source system quickly – as well as accuracy. Most likely, companies with mature IT departments already execute some sort of Operational BI, perhaps without realizing it.
The primary characteristic of usage is that frequently users report against source systems (either directly or through a staging area), and often they are interested in getting only a limited subset of data, such as focusing on a single client, store, or stock.
IT staff required to build and maintain the solution includes DBAs, ETL developers, data modelers/architects, as well as SME in the software system of your choosing. If you decide to custom-build one, you would need additional developers and tech leads. Close coordination with infrastructure folks to define such issues as security and latency would also be needed.
On the business side, you would need experienced business analysts who could properly identify and define business processes, producing usable requirements for an IT implementation team. In addition, they would be responsible for maintaining the process map, ensuring that any change in business rules gets communicated to the IT team.
One should expect most of the source data to originate from enterprise-wide ERM or CRM systems. However, it is also likely that certain elements would arrive from smaller-scale systems, such as e-commerce platform or web-based tools. Note, too, that reporting from multiple sources in real-time fashion could be challenging depending on the complexity of integration activities required.
In the following post, we will review “Traditional BI” approach.