With social bookmarking and tagging becoming popular across most consumer websites, such as del.icio.us, flickr, youtube and others, are taxonomies dead?

Emphatically: no.  Every organization that generates or manages volumes of information should have and maintain a commonly shared taxonomy for organizing, storing and accessing that information.  Just so we’re clear: when I say “taxonomy”, I’m not talking about stuffing dead animals.  That’s taxidermy, and I’ll save that for another post… Maybe.

A taxonomy is simply a set of commonly used terms to classify a piece of information.  In practical terms: a corporate taxonomy is a hierarchy of related terms used to organize documents and information used across the entire company.  An example of a single branch within a corporate taxonomy would be something like: Product Line X > Product Y > Product Y Feature Z

So what are the benefits of a corporate taxonomy, and where do you start?  Let’s take a look.

The Benefits of a Corporate Taxonomy

Before I go into some suggestions on how to build a corporate taxonomy, let’s examine why.

• Spend less time looking and more time discovering documents and information. Arm your sales force with the right information to close more deals. Bring projects in on-time and under budget because you’re able to locate the how-to when you need it.  Make cross-selling a reality instead of a pipe dream.
• Spend less time publishing and more time discovering documents and information. Having a common hierarchy and a process for determining where documents get published makes it easier for your knowledge workers to know where to publish their documents and information.  Some of this can be automated as well, but having a broadly understood classification of information prevents confusion and lowers the likelihood of documents and content being published incorrectly.
• Unify the vocabulary of your company or organization. Often times if a company or organization lacks a company taxonomy, they may also not even agree on a common vocabulary.  This can especially lead to confusion and inefficiencies when disparate groups from throughout the organization attempt to communicate with each other.
• End-user tagging still generates random terms at best. In a company setting, it may not even be possible to engage all users to tag information, but worse than that, critical information can end up being categorized incorrectly.

Building a Corporate Taxonomy

Starting a taxonomy from scratch can be challenging, especially if you don’t already have a commonly used set of terms across the enterprise.  This is by no means a comprehensive list of the steps involved, but rather more of a pointer to get you started in the right direction.

• Get corporate sponsorship before you begin. Effectively building a corporate taxonomy requires that you have people at the right level of the organization who can support your final decisions.
• Identify the key stakeholders and keep them constantly informed. This is pretty much a generic requirement for any major project.  Are there people throughout your organization that will be significantly affected by your decisions?  Keep them informed, and set expectations as to how you will communicate the final outcome.
• Poll your consumers for the terms they know and use today. It’s likely that large groups of users already use common terms for the various classes of information they generate today.
• Try to keep the taxonomy topical and avoid document classifications in the structure. The types of information available typically can be understood by looking at the individual content item itself rather than navigating to it through the taxonomy.
• Try to avoid having more than three layers of depth. When information is over-classified, browsing for it and finding related information can become difficult, if not impossible.  People lose faith in the taxonomy and may be unwilling to use it.
• Present a taxonomy skeleton to key members of your organization. Ultimately it’s up to the business stakeholders and your sponsors to sign-off on the final version, but don’t be afraid to assert a clear recommendation where it will clearly benefit the structure of the taxonomy.
• Make sure there are no empty branches. Once you ultimately define your taxonomy, make sure that there are no empty branches, or branches that only have one or two pieces of information.  It may be impossible to avoid this, but having a lot of empty folders may ultimately lead to a lack of confidence in the overall organization, or it could be a sign that there are too many layers of classification.

Tagging in the Enterprise

If you’ve read this far, you may believe that I’m underestimating or even dismissing the power of tagging.  Indeed, I do believe that tagging is essential to adding a further layer of classification over content and information, and that it can add additional value.  However, imagine for a moment a public library that decided not to conform to the library of congress standards for cataloging their books.  Could we expect them to locate the books we need if they didn’t agree on a common classification system?  It would seem unlikely.

In fact, products available on the market today, such as librarything.com, offer services for libraries to extend their catalog systems with user-defined tags.  Amazon.com follows a similar structure, in that they allow users to tag products, but ultimately they maintain the overall structure. This is the right hybrid model: a common, reliable scheme for classifying books and products, with user-extensible metadata that can supplement the taxonomy.

Think I forgot to mention something?  Care to disagree?  Let me know in the comments!

Filed under: Consumer Web, Enterprise Web, Knowledge Management