The main constraint the business user experience, is that the data needs to be present in the system, in order for them to perform analysis on it, and on a day-to-day basis the type of decisions business users make, often requires information that is not yet in the data warehouse and perhaps never will be. Self-service business intelligence tools have in recent years become more and more widespread to enable just that. It shifts the focus away from the processes required to manage data in a central data warehouse and towards managing and utilizing data as needed with minimal involvement from IT departments.
Microsoft Self-Service BI
Self-service business intelligence gives the user, the power to build their own data models, reports and analysis, often mixing in data that is not currently present in the data warehouse, either because it does not make sense there, or simply because a new need for this particular data has risen. Gartner refers to self-service business intelligence as “faster, more user-friendly and more relevant BI”*.
Microsoft addresses self-service BI in the latest addition to their stack of BI tools, and while still in preview official prices and release details have been revealed. Power BI is a number of integrated components, tightly integrated with Office 365:
However, with the introduction of self-service BI, the organization and in particular the IT department, faces a number of challenges. Enabling business users to create and share their own reports, and in the process, accessing and perhaps building their own data sources and data models, self-service BI can potentially complicate data governance and skew the notion of “one truth”, which has always been the main driving force behind building and maintaining a central data warehouse. Without proper governance the business risks include:
There is no clear and “one-fits-all” recipe for governing self-service BI, but a few key areas must be addressed in the approach of a successful implementation of self-service BI within the organization.
One of the most important distinctions to make when introducing self-service BI into the organization is to identify the target audience and divide them into the appropriate user group. Most often, business users are satisfied with a set of standardized reports and dashboards, and only a few requires the need to perform analysis in an ad-hoc “data discovery” fashion. According to the survey performed by Wayne Eckerson from Inside Analysis, one user responded, “Self-service BI is great for users with analytical experience, but bad for users without an analytical background”.
Data quality and security
With self-service BI, data governance becomes especially important, because of the introduction of perhaps ungoverned data sources and the freedom for the business users to create their own data models. Ongoing data quality checks and data security processes must be identified to ensure that the data used is accurate, up-to-date and secure. The role and assistance of the IT department is especially important in helping the business users understand where data quality issues exist or can arise. Appropriate standards for accessing data sources and maintaining a high level of security throughout the business is key, and at the same time appointing data stewards within the business to ensure that these sources are up to date and safe.
Once self-service BI has been introduced, if appropriately monitored and governed, it can serve as a great guideline for the planned development of the future central information strategy within the business.
Microsoft has introduced a management dashboard for monitoring the use of self-service BI, primarily centered on the use and execution of Power Pivot models.
By monitoring the dashboard, the IT department, it can gain insight into information on server health, workbook activity and data refresh within the workbooks.
Self Service BI is still very much in its early adoption stages among many organizations. It bridges a gap that has long been growing, as enterprise data warehouses become ever more complex, and it provides, if governed correctly, a great balance between the analytical business user that want the freedom to access, analyze and create their own data models, and the casual business user that simply would like access to the information without having to much interaction with the data.