People love their comfort zones. That’s what makes it so difficult to “teach an old dog a new trick.” If you ask managers to leave their Excel spreadsheets behind and do all their analysis on a reporting tool, you can almost see the blood running out of their veins.
Investing in self-service business intelligence (BI) only succeeds if you take time to make it truly work for business users.
Winning Over Business Users with Self-Service BI
After spending countless hours teaching business users how to use a self-service BI tool by simply dragging and dropping the objects that they need, quite often, they’ll ask you how to export the query result into an Excel model that they built for analysis. By doing this, business users are costing their employers millions of dollars for simple data extraction when they could use common database management tools for only a few hundred dollars or simply write a few lines of VBA codes that are available in Excel at no additional cost. So what’s wrong with this picture?
It’s not impossible to change the mindset of business users, but don’t expect them to jump out of their chairs to embrace self-service BI wholeheartedly right away. And don’t expect to simply dump the tools on business users’ laps and have them start using them on their own, which is what self-service might falsely imply. It takes effort to teach business users the true value of self-service.
Typically, a data warehouse isn’t built for self-service BI. It’s worse when data is stored in a recursive fashion. There’s no way that business users—including power users—know how to write complex Structured Query Language (SQL) to query the database. In order to make self-service a reality, you need to:
- Properly rearrange the data in a reporting data mart that allows the data to be easily presented in a report format
- Shield users from the technical complexity of SQL by setting up metadata layer
- Provide plain business terminologies instead of the typical database gibberish
So what should self-service BI be? The term self-service is misleading. Perhaps end-user BI would be more appropriate. Name aside, self-service BI should be a close partnership between IT and business users. IT creates and maintains the environment with guidance and feedback from business users, and then works hand-in-hand with business users to perform business analytics: slice and dice the data, build what-if scenarios, and establish trending over time.
Bottom line—self-service BI isn’t a different type of BI; it’s what business intelligence should be: IT and business users working together to create more powerful insights that business users can easily access and understand. When organizations do this properly, they reach true BI enlightenment.