Recently, I read an article from CIO.com examining the 5 pitfalls of self-service BI. The author, Thor Olavsrud did a fantastic job explaining how just having a tool doesn’t mean that people will use it properly. “Garbage-in, garbage-out” is a phrase I’ve heard many times in my career, and I think it applies to self-serve BI in many different ways.
Olavsrud identified several issues, including:
- Business metric chaos arising from smaller business units defining their own terms (something I’ve personally experienced multiple times in past roles)
- Turning [inexperienced] users into data engineers
- Data security issues
- Scalability issues
- Cost (especially if the company loses control over the number of licenses in use)
Bernard Marr made similar observations in his Forbes post, Why We Must Rethink Self-Service BI, Analytics And Reporting. In it, he points out the risk to an organization that does not centralize interpretation and analysis of its data. Not only could someone buried in a department miss out on the big picture, but without formal training they might even be drawing the wrong conclusions in the first place.
At 3AG we took a similar look at this situation in our, Ad hoc analysis: business intelligence’s fast food problem. Self-serve BI has some big advantages, such as being easy to use, accessible to all, and flexible. However, the advantages disappear when ad hoc reports become permanent fixtures within an organization. Further, as noted by Marr, results developed by data “novices” risk misinterpretation. And if a company hasn’t properly prepared its data and supporting infrastructure, the results are probably not even worth looking at.
There are no shortcuts in both life and business. When it comes to data analytics, companies should still invest effort in building proper infrastructure, maintaining good data hygiene, and most importantly, ensuring that the insights they share with their employees are accurate and insightful.
