You talk and blog a lot about BI best practices and how to adopt them. What’s a BI problem or worst practice you’ve seen consistently that doesn’t seem to go away?
Overloading data – hands down. I still can’t believe how much time companies spend on extract, transform and load (ETL) processes at the expense of deploying user functionality. After all, the data model is done – isn’t it time to populate tables? This is when we have the “we love you, but we have to kill you” conversation with our clients.
It’s so much easier just to load all of the data. Except for two problems. One, you’ll never load all of the data as long as new requirements keep cropping up. (And I hate to break the news, but if new requirements don’t keep cropping up, you’ve got bigger issues.) And two, your business users are thinking, “They’re still loading data? What about my scorecard/portal/dashboard/reports? They must not understand my business.”
Fair? Probably not. But perception is reality, and sometimes reality bites. Chunk up slices of functionality and load the data proportionally. Of course there’s going to be some rework! But the price of data loading in a vacuum is disaffected constituents. And that reality bites even harder.
Dig Deeper on Business intelligence best practices
Related Q&A from Jill Dyché
There’s a lot of talk about master data management but not so much about CDI and PIM. Find out what happened to those MDM-related technologies. Continue Reading
Thinking about adding mobile business intelligence (BI) tools to your BI and analytics strategy? Find out where the real value of mobile BI resides ... Continue Reading
Learn about the most common master data management (MDM) project pitfalls that companies run into. Get a list of the major problems that can hold ... Continue Reading