Where is the excitement around this issue?
It seems that “Taxonomy” was my word for the week. This is my third post about it within 7 days. It’s not that I am in love with the word, rather, it’s just pretty darn important! With any big initiative, the first thing we look to is a solid foundation for communication. Think about it, we usually address taxonomy anywhere from casual discussions to formal governance groups for many initiatives – dare I say any initiative that strives to bring real change to an organization begins with taxonomy (either consciously or subconsciously). Thinking it through, here are some of my top-of-mind game changers that require a solid taxonomy:
Master data management. By definition this is really an enterprise taxonomy that is the official reference data for an organization.
Metadata management. Tagging data with information is best performed only after a taxonomy is well-established. Else, with-what-shall-I-tag-it plagues the process.
Business Intelligence. Without a proper taxonomy, how do we bring together people from diverse business perspectives together to understand data from a central and enterprise viewpoint?
Data Management. Well, of course we can’t properly manage data without knowing where things are in a hierarchy or what context the data should exist in.
Data Quality. Here we are really measuring data against the taxonomy; whether implicit or explicit.
Governance. Strategic decisions are made for specific purposes and they need to rely and depend upon a socialized and accepted taxonomy.
Data Stewardship. This is the process of holding someone accountable for making tactical decisions to implement strategic direction.
BPM. When we look to manage business processes, they depend upon real information. So, having a taxonomy to base these data points is crucial.
SOA. Reusing software components and exposing them at the enterprise level demands a highly accepted understanding of the organizations data. Sure, this view is exposed as a group of web services that are published in a repository that is self aware – but without a canonical data model as your underlying foundation, consistency is not reached. A canonical data model is highly correlated to a mature taxonomy.
Strategy, Solutions and Architecture. It’s near impossible to calibrate these three without a friction free flow of communication. Let’s not talk about what should’a, could’a and would’a – but let’s focus on the business problem at hand. Having a living taxonomy that is socialized, accepted and part of our DNA is key to gaining quick momentum as we put distance between us and our competition.
These are just some quickies that I bubbled up. What other initiatives need a solid taxonomy? Thinking about taxonomy, when you look to bring real change to an organization, what happens? From my experience, there are two choices:
1. Address taxonomy early and often. Realize that there are some things that are so important that we need to establish, socialize and enforce them.
2. Jump to build a solution. Then realizing there are terms misaligned, misused and duplicated, go back and either fix the data models (and all subsequent diagrams and code – this rarely gets done) or create a lot of code to hide these issues. When we do this, we establish a short term brittle foundation that breaks when the next change comes or we end up with a bunch of custom spaghetti that tries to tie things together but really ends with just a lot of confusion.