When choosing what direction to pursue in the data warehouse (DW)/business intelligence (BI) application, there are two distinct interrelated parts. The DW portion has its own considerations, but the structure and accessibility of the DW is heavily influenced by the BI needs. After all, the BI portion of the application is the front-facing success measurement of the application.
Let us assume that technological issues have been considered; these include database software and structure, source application systems, third-party data feeds, ETL tools, data model structure, and data access. It almost goes without saying that any BI tool must be compatible not only with the DW application but also the user portal environment(s). These considerations apply whether an DW/BI application already exists or is merely a contemplated portfolio expansion.
Identify how the BI tool is going to be utilized in the organization. Will this be the “official” organization data (the “single source of truth”) or does it supplement other applications the way annual statements support statements.
Determine the user audience. No doubt there will be occasional, casual, and power users. There may be third-party users to factor in, such as independent agents. There certainly are categories of users within the organization’s respective departments. Front-line underwriters probably do not have the same needs as their supervisors or managers. Establish the needs of the respective audience. These identifications will determine both data (information) exposures as well as assist in defining the type of BI tools to deploy.
Determine what type of reports each user classification can access. There may be a group where only pre-generated traditional reports are needed while another group requires pre-generated reports supplemented by some ad hoc capabilities. And there will always be power users (think actuaries) whose needs require robust ad hoc report creation capabilities.
Where do dashboards and other visual capabilities fit in? The old saying a picture is worth a thousand words is certainly applicable when considering a BI tool. What capabilities are required? What are the user license structures? Are the platform’s users going to be occasional/casual users, power users, third-party users, or developers? Where does data security and application access fit in?
Selecting the appropriate BI solution can be a stressful process. Cost, of course, can be a deciding factor. With scores of software solutions available in the marketplace, exploring the available software landscape can be a daunting task. For help navigating the BI solution marketplace, read Novarica’s report Business Intelligence Solutions for Insurers, which provides an overview of current solutions in the US marketplace.
Rory Read is a former Vice President of IT at Columbia Insurance Group, and a current Novarica Research Council Fellow. He can be reached at [email protected]
Add new comment