Success Factors in Building an Enterprise Taxonomy
A few months ago, we had the pleasure of responding to an RFI that not only had the typical methodology and vendor information requests but also had a number of really thoughtful questions aimed at sussing out expertise and philosophy. Taxonomy can be more of an art than a science in a lot of cases and its important to find a taxonomist that you feel gets your situation and has the same taste in “art” as you do.
One or two of these questions forced us to really sum up some of the key success factors that affect a corporate taxonomy project.
They were:
- What Corporate Metadata Model and Taxonomy “Best Practices” should be considered?
- What “approaches” do you recommend for implementing a Corporate Metadata Model and Taxonomy (including brokering common language, helping different communities communicate with each other without giving up their identity, supporting our discovery processes, etc.)?
Taking both of these together, we can say that successful taxonomy and metadata projects depend on a number of key factors and best practices:
- How well stakeholders and core users understand and buy into the concepts of taxonomy and metadata, and are engaged throughout the process Education and stakeholder engagement are instrumental in any taxonomy project. Making metadata and taxonomy implementation successful in any system or process design requires a core group of people who really “get it” and are willing to stand behind requirements and design recommendations. Users who will be asked to apply metadata must understand the benefits or functional imperatives behind the use of shared language before they will be willing to comply.We recommend making stakeholder education a priority, starting early and continuing throughout the life of the project to create a baseline of understanding. We also advocate involving system users in the design process as much as possible to ensure a sense of ownership and feeling that any taxonomy and metadata structures have been built “by us, for us”, rather than imposed. Involving users in activities such as card sorting, requirements gathering and review helps create this sense of engagement that can last well beyond the end of the project.
- How well the taxonomy and metadata is integrated into the overall information management strategy and content lifecycle A taxonomy and metadata strategy must fit into and support the organization’s overall plans for information management. To make sure that all the right stakeholders are involved early enough in the process to help guide direction, you have to have a sense of what the taxonomy’s role will be in enterprise standards, systems and processes.From a technological perspective, an evaluation on how in-flight and planned IT initiatives affect the scope of the taxonomy is necessary – how it will be implemented in future systems or interact with legacy systems. From a process perspective, there must be a vision for how the taxonomy will be factored into such enterprise initiatives such as records management and collaboration.
- How pragmatic the taxonomy and metadata models are, in relation to functional needs and technological constraints/opportunities
Users have limited time and interest for metadata and taxonomy, so we recommend developing structures that limit the amount of effort required to organize content, relying on automated processes where possible and restricting the framework to what is imminently practical and brings valuable functionality. We believe in creating metadata for 3 distinct purposes only:
To support a business process/function
To support search & retrieval (findability)
To support records management - How the taxonomy helps bridge gaps between technologies, languages and people without compromising functional need Taxonomies are intended to create a shared language that helps reduce barriers between systems and people to increase findability & usability of knowledge. This is a beautiful thought – however, it isn’t always possible to create a common language due to differences in culture, process or technological system limitations. Information is often application-centric, departments and processes are often fragmented. It’s important to identify these differences and work around them using cross-mapping (a.k.a crosswalks, translation layers) between different vocabularies where needed, rather than imposing an approach on a tool or group that does not fit their context.These crosswalks can be invisible to end users and create an enterprise-wide view at the right level. In the end, you have a set of corporate shared elements and a subset of vocabularies that are more system- or group-specific and mapped back to a common language.
- How realistic and incremental implementation is – mix of top-down and bottom-up approaches All of the above being said, big semantic technology/taxonomy/metadata projects often get bogged down in complexity. Tackling a corporate-wide, holistic taxonomy & metadata model can sometimes take too long and try to incorporate too many applications at once. Business users will often get impatient and go off on their own, creating their own niche vocabularies and structures instead of waiting for the corporate structure to be complete.We recommend mixing a top-down and bottom-up approach.Top-down attention should be paid to create an enterprise standard/framework for core metadata and key shared vocabularies that are common to multiple applications and user groups. This helps set a certain level of enterprise control over base elements involved in wider initiatives, such as records management.At the same time, a more bottom-up approach should be taken for more granular and extended taxonomy/metadata implementation.Focusing on a specific business process or system can help define and restrict the scope of implementation and allow the project to show meaningful and relatively quick success. This helps build support for the taxonomy and champion it to other groups that can continue extending and adapting the work.
During development, many additional specific taxonomy & metadata best practices are followed concerning vocabulary structure, format, encoding standards, etc., depending on the context of use. These would be outlined during the assessment of any existing vocabularies and taxonomy/metadata stakeholder education. If you want more information about the role of enterprise taxonomy, feel free to drop us a line!
Image credit: http://www.totalhrmanagement.com/employees-human-resources-and-social-media-best-practices-top-questions-to-consider