In the Information Age, you might expect taxonomy to be a fairly well known term. Certainly, there are “academic” definitions for taxonomy, just as there are published standards describing taxonomy concepts. Taxonomies are growing in use but the definitions used have always been murky. For years, I would get frustrated by people’s “misuse” of terms like “Taxonomy” , “Ontology”, “Lexicon”, and “Thesaurus”.
Then, I just let it go.
As someone who has had job titles like “Chief Taxonomist” in the past, you’d think I’d be a bit more of a stickler for definitions. But I am not. The reason I gave up worrying about the labels is that they just didn’t matter in this case. In “Letting Go of the Words”, Ginny Redish wrote that effective web content has to be a part of a conversation. At the end of the day, taxonomies are a means to an end and are rarely valuable on their own. They’re the bridge in the conversation between a business and its customers, an information system and its users.
A good taxonomy is one that is designed to solve problems. If business problems can be solved with an ANSI compliant thesaurus, all the better, because standards are a good thing. But if the business problems are solved by a different structure leaning towards ontologies or just a simple list of terms, then the best approach is to follow the business goals and requirements. The labels that matter are the ones with which end users interact, and those are not necessarily the same ones driving the implementation. Ultimately, successful taxonomies meet the needs of the intended users. An external facing taxonomy will jettison internal naming conventions and dance between user expectations and the organization’s business goals.
For example, we were working on a taxonomy for a major specialty retailer with significant e-commerce and brick-and-mortar presence. They wanted us to look closely at their taxonomy and help them shift it to better support evolving branding across multiple digital touchpoints. As their current taxonomy was fairly large, the assumption was that a different – but equally large and complicated – taxonomy was necessary. After a thorough review of the new branding goals, user research, analysis of sales and web analytics, and assessment of their internal processes, it turned out that what they really needed was about ten concepts applied consistently across different types of information in multiple backend systems.
Technically, this list of ten concepts could be considered a very simple taxonomy, but really, it’s a core description of the company’s interaction with its customers. For the company to interact with the customer in this way, these ten concepts needed to be associated with products, customer history, stores, activities, and most of the other customer facing assets. So, is this a taxonomy? A controlled value list? An upper level ontology?
The customer didn’t really care about a technical definition, using any of those terms when talking to company executives would’ve probably killed the project right there. In this case, we referred to these as brand pillars. Executives understood those terms and the project got funding.
Freeing ourselves from the taxonomy nomenclature made it easier to work with different groups within the organization. The marketing team loved the notion of “brand pillars” and then understood the need for integration, governance, maintenance, etc. The IT and MDM teams smirked at “brand pillars,” but were right there with us when we talked about a simple look-up table to be used by different databases and shared across systems. The business owners were just pleased that someone was listening to them. These discussions were possible, not because we were discussing taxonomies, but because we were talking about supporting the omni-channel challenges the company was having.
Following the business goals and the culture of an organization is the key to getting a taxonomy project to move forward. Ultimately, a taxonomy supports core interactions between people and information. In projects, getting to the business solution is often more interesting than nailing down a definition.
An excellent discussion of the standards and types of taxonomies can be found on Heather Hedden’s site.