Most people know "The Story of the Three Bears." While ostensibly meant for children, I have found that it's an important parable for organizations with respect to data layers. Let me explain.
Some organizations place far too many layers between end users and enterprise data. I'm talking about the usual suspects: dual data entry, the lack of a master record, redundant or superfluous systems, the lack of real- or near-real-time data, and out-of-sync datamarts and data warehouses.
I've personally worked with more than a few of these enterprises. The end result was typically data anarchy. No one could determine accurate lists of customers, employees, products, etc. Decisions were suboptimal. Opportunities were squandered. Dysfunction prevailed.
At the other end of the spectrum, there are organizations that in my view don't place enough barriers between end users and enterprise data. (Note that these are typically small companies.) I'm talking here about perniciously giving employees the ability to directly add, change and delete data stored in databases. This usually causes a number of important issues:
- Circumventing business rules reverberates throughout other parts of the system, causing unforeseen problems in the process
- Best data governance practices are ignored
- Potential regulatory violations, especially SOX
- Data anarchy
Simon Says: Aim for Just Right
Not often, but every once in a while I encounter a thoughtful organization that establishes appropriate layers separating everyday end users and organizational data. There's no one right number of layers here: it's typically greater than one and fewer than four. "Just right" organizations strike the right balance among legitimate organizational controls and the need for employees to do their jobs in an unencumbered manner.
That should be the goal: as many layers as you need – no more, no fewer.
What say you?