Darin Stewart

I’ve been playing with a new core feature of Microsoft’s SharePoint 2010, the Managed Metadata Service.  The MMS has at its heart the idea of managing terms and keywords and applying them to documents. So, metadata and taxonomy. Now, were I a product manager or marketing executive for Microsoft, I would be hard pressed to come up with two less sexy features to promote a new product. So why is Microsoft putting metadata and especially vocabularies front and center in SharePoint 2010?

My contention is that Microsoft has finally realized or at least acknowledged what the rest of us have known all along. Metadata isn’t a “nice to have.” It’s a “must have.” It is the enabling factor behind successful content management and collaboration. This isn’t new, but what might be a new idea for some is that while metadata is the enabling factor behind ECM, managed vocabularies (to avoid the “T” word) are the enabling factor behind metadata.

So first, what exactly is metadata? We all know the standard trope that its “data about data” or “information about information.”  While that definition might be sufficient to get us through a conversation at a cocktail party, we need to go a bit deeper if we want to take full advantage of metadata in the enterprise. One of the best general purpose definitions of metadata I’ve encountered comes from Mike Blechar, a Gartner Research VP and Distinguished Analyst.

Metadata is information regarding the characteristics of any artifact, such as its name, location, perceived importance, quality or value to the enterprise, and its relationships to other artifacts that the enterprise has deemed worth managing.

It’s the part about what we “deem worth managing” that is important. Whenever we want to manage something, whether its people, content or widgets, there are certain tasks that need to be performed. The more we can automate or at least simplify those tasks, the better we will be able to manage those things we are responsible for. That’s the role of metadata. It enables us to simplify, standardize and in many cases automate the management of our content. Without that extra bit of information, we really haven’t moved very far beyond sticking hardcopy in physical filing cabinets. We still have to perform all of our managerial tasks more or less manually.

The more thought and effort we put into our metadata the more we can do. This requires us to think about metadata beyond simple keywords and file types. Different types of metadata enable different capabilities. We are most familiar with descriptive metadata that tells us what the content is about, who created it when and so forth. This is arguably the most important, because it is what helps us find the content we need. But other more administrative data, such as access permissions and retention policies can actually facilitate some pretty interesting things like automatic storage and routing management, access control and multi-stage disposition, to name a few.

If you’ve already drank the Kool-Aid and attempted to undertake a significant metadata effort, all sorts of feelings of disillusionment, cynicism and outright bitterness may be bubbling to the surface right now. Because, frankly most metadata initiatives fail. This is true even if all the metadata gets defined and the content gets tagged. It just never quite seems to live up to the promise. This is for several reasons. First, defining, developing and deploying metadata is a lot of work and requires a specific skillset that often does not exist in the enterprise. In order to be useful, our metadata needs to be integrated with the tools that can make use of it, our search engines our web content management platforms, our BI tools, and that takes a lot of planning and a lot of effort. The other thing that people tend to forget, especially sponsors, is that a metadata initiative is never “done.” In order to remain useful metadata must be consistently applied, managed and reviewed. Field definitions need to be updated as practices and importantly, as terminology shifts over time. This last point is critical. In my view, the real reason most metadata never realizes its full potential is this lack of attention to the values we use to populate our metadata.

Think about all those keywords you painstaking add to documents just so people will know what the document is about and so they can find it when they need it. If you and I aren’t using the same terminology or labels to describe a particular concept, it’s very unlikely that you are going find what I produce even if I do tag it because you will be hunting for a different term. We all have our own vernaculars and we tend to be wildly inconsistent when using those internal dialects. Again, if the vocabulary in searcher’s head doesn’t align with that in taggers head, it is unlikely that they will find the information they are looking for or that they will be able to correctly interpret it even if they do.

For example, if you tell a colleague “my office is on the first floor” someone from the UK is going to head for the stairs while an American is going to be looking for you on the ground floor.  As another example, in America, the title “Director” more or less means a glorified manager. Again, in the UK it means something altogether different. It is a much more prestigious and powerful position. So you can see that once we get outside of our own environment, its easy for things to get confused. This is the problem Managed Vocabularies are intended to address.

Managed vocabularies are gaining favor in the enterprise, because they are useful. The tree structure, the parent/child relationships traditionally used to organize these vocabularies are a powerful if under appreciated feature. They can serve to guide the user through an unfamiliar information space and demonstrate relationships that might otherwise be missed. It goes a long way to bring some order to any given knowledge domain which otherwise could be quite chaotic.  The new Managed Metadata Service is a huge step forward for metadata in SharePoint and can begin to introduce useful tagging to the enterprise audience, but only if done thoughtfully, starting with a few critical, broadly applicable elements and most importantly, by adopting a few well crafted managed vocabularies to populate them.

2 Comments »

Category: Collaboration Enterprise Content Managment     Tags: managed metadata service, managed vocabularies, Metadata, mms, sharepoint, Taxonomy

Share

Tweet