In the last 18 months, I’ve seen real growth in the number of enterprise taxonomy projects we’ve been doing for clients. And despite the many benefits a strong enterprise taxonomy can deliver to an organization, I see three challenges organizations face when trying to use them effectively:
- Taxonomies can be static. They’re represented in “flat” media like Excel spreadsheets and Visio charts.
- Taxonomies can be very manual. They require lots of effort to implement them within systems, as well as to keep them current.
- Taxonomies can be abstract exercises. They live only in deliverables, not instantiated in an organization’s systems and processes.
However, I’m working on an enterprise taxonomy effort now that’s looking to be quite different from the typical engagement in a number of important ways and is on track to overcome all these challenges.
Ordinarily when I do an enterprise taxonomy, it’s a foundational project for the organization: They’ve never done taxonomy work in a sustained, thoughtful way before, or if they have, they’ve never been able to extend a single taxonomy across the enterprise. So we’re taking baby steps, and one of the biggest wins of this kind of project is just to make folks aware of what taxonomy is and how it meaningfully impacts their business.
In this case, however, the organization has very mature enterprise architecture and information management capabilities, so they’ve accomplished a fair bit of taxonomy-related work over the last couple of years. Most importantly, they’ve addressed head-on how they create, store, and retrieve the information in their transactional systems, reporting engines, and other managed and structured applications (i.e. their structured data). And although they don’t have a true enterprise taxonomy that they can use to classify their documents (i.e. their unstructured data), they do have a mature business process architecture built out, which is a great place to begin addressing document classification.
They also have some very robust enterprise tools in place that can be used to help manage their taxonomy—IBM’s Rational Asset Manager (RAM), both Enterprise Meta Environment (EME) and Enterprise Metadata Management System (EMMS) from Ab Initio, and Envison—and they use them. This is in stark contrast to almost all the organizations I’ve worked with, which either have no tools in place for managing taxonomy/metadata, or they have tools they don’t really use in a meaningful way.
Enough context. Let’s look at how they’re poised to address the three challenges I cited above.
Challenge 1: Taxonomies Can Be Static
Right out of the gate, this organization is positioned to avoid the first challenge, that taxonomies can be static. Given the tools they have, they’ll be able to load the detailed results of the taxonomy project (ordinarily a spreadsheet containing the parent-child relations between categories) into a relational database that can be accessed by RAM, EME, or EMMS. From there, the taxonomy can be coupled with the metadata standards they already have in place, represented in metadata fields that can be syndicated throughout their structured data repositories, associated with the assets managed in RAM, and so on.
And with the process modeling capabilities they have through Envision, they’re able to free the taxonomy from the bounds of flatland: by clicking on categories in the taxonomy, users can move up and down the parent-child hierarchy or see the documents and metadata associated with each category—a far cry from flipping through a combination of 11×17 PDF pages and spreadsheet tabs to navigate the taxonomy.
So far, so good. In the next post, we’ll take a look at the remaining two challenges and how this client is meeting them.