The main strength of the Enterprise Information Flow approach to information management rests in the plethora of analytical possibilities it provides. We’ve suggested the fact, a little, in our previous pieces about data quality and data flow, but today we are going to present it in its entirety.
What kind of analyses can be based on the comprehensive implementation of Enterprise Information Flow?
- Impact Analyses. Before making any change on any piece of code (attribute, dataset, etc.), it is necessary to know how it will affect the whole system. For example, a change in a data structure could generate a change in the calculation demands of new datasets.
- Data Flow Analyses. Every element’s origin needs to be fully searchable. There are basically two types of sources: direct (datasets and records the element was derived from) and indirect (other sources somehow affecting the element’s sources). Of course, every transformation also needs to be noted and traceable.
- Information Flow Analyses. Data Flow Analyses work on the physical (technological) level of data transformation; Information Flow Analyses focus on a higher level of already transformed business information, but their function is the same.
- Data Attribute Analyses. EIF can work with a lot of different data attributes. The analysis of values and behavior of those attributes in data and informational flows can differ greatly. Let’s take a look at some sample questions you might ask yourself: How many people are responsible for, connected to, or creating final output? Which technologies work with sensitive data? Is a transformation affecting data security? Which technologies are connected to a specific part of the system? Are reports free of sensitive data?
- What-if Analyses. And, last but not least, what-if analyses try to determine what changes would occur in information flow if some things were changed. For example, what would happen if you changed the ETL tool, Master Data Management solution, some of the production systems, etc.
Those are the basic five types of analyses you will be performing (and likely already are or want to) in your information management system. Enterprise Information Flow integrates all five of them and, if implemented as a whole, will help you to maintain control over your system, its resources and outputs.
Is there anything we have missed? Is there a type of analysis you need to perform which is not listed above? Share your opinions with us in this Reddit, follow us on Twitter or simply write us an email.