It Starts With A Question
It all begins innocently enough: the end-user makes a request to pull a collection of data out of your enterprise resource planning (ERP) system to help answer some ad hoc questions, or provide information to outside sources.
Everything works splendidly, the data is provided in a spreadsheet. The end-user does a quick sort here, a little filter there, maybe a couple of computed columns derived from the original data to help make it easier to present. Questions are answered. The end-user is happy. Life is good!
Then It Happens
Weeks later, some less-than-desirable enrollment numbers get presented in a departmental meeting and someone asks, “Where did that come from?”
“Why, it came from our ERP system, of course.”
“Well, then it must be right!”
Not so fast!
Yes, the data originated from the ERP system, but that was weeks ago. Since then, thousands of students have registered, added or dropped classes, changed majors, applied for graduation, and generally acted like students.
What started innocently as a helpful way to answer questions can become a nightmare when the misperception spreads across campus that enrollment figures are dreadfully low! Will there be budget cuts? Staff reductions? The sky is falling!
Beware The Shadow
This scenario demonstrates the risks of a well-intentioned data dump: Every data dump performed creates the potential for a “shadow” database to be born. Why should you worry about shadows? Consider:
- The “shadow” data is no longer sourced from the ERP system, the authoritative system of record for the institution.
- The data was only good for that moment in time and was likely created for a specific purpose that may not be the same the next time someone pulls data from it.
- Data dumps are often raw and unprocessed. Once the data is separated from the processes that put it in place, it can be manipulated, misinterpreted, misleading, or meaningless.
- The data may include unmasked, personally identifiable information.
- The data is no longer as secure as when it resided in the ERP system. It can now be transported easily via email or USB flash drive, or stored on a poorly secured laptop computer.
While there may be times when it is appropriate to provide a data dump in response to a simple question, care must be taken to ensure the data is used only for its specific purpose within the proper time frame, remains well secured, and is properly labeled.
How should requests for such personal data dumps be handled? It’s always preferable to rely on your enterprise reporting solution for accurate reports and dashboards from authoritative data sources, including but not limited to:
- ERP System
- Operational Data Store (ODS)
- Enterprise Data Warehouse (EDW)
- Institutional Data Mart
Use reports and dashboards to provide the right data at the right time and context, from the right source. By doing so, you can provide meaningful results that include:
- appropriate data elements
- accurate, verifiable and auditable data
- relevance for the target audience
- compliance with regulations and standards
- consistency over time
- documented report description and notes
- adherence to institutional data standards
- freedom from human intervention or manipulation
- automatic delivery
- flexibility to display or print graphically
- ability to analyze using multidimensional online analytical processing (OLAP)
Ultimately, it’s about providing the best solutions for our end-users. We all want to provide solutions that can be presented with confidence. The next time a data request comes to you, what solution will you choose: data dump or enterprise reporting solution?