When a taxonomy design is done well, it goes unnoticed, as it enables a seamless connection between the end user and the content. However, as business needs evolve, it’s inevitable that a taxonomy design must also evolve in order to maintain this seamless experience. Without appropriate taxonomy governance, taxonomy designs can quickly fall into chaos, resulting in poor usability and findability.

If a taxonomy is poorly managed, an organization will have little choice but to embark on a costly taxonomy redesign/refresh initiative. The price of these vary, but can easily cost in the range of $60k-$100k. The prospect of investing in such a project every few years quickly compounds.

Investing in taxonomy governance is a critical approach to avoiding these undue costs and continuously evolving a taxonomy design for maximum value and effect. In contrast to a complete taxonomy refresh, robust governance can be achieved without a large investment in support and management. Governance simply entails creating and adhering to roles and procedures that allow for long-term ownership and responsibility. These ensure that the evolution of a taxonomy happens in a managed, predictable way. At EK, we refer to this as sustainable evolution of a taxonomy design. Below are several best practices for creating effective taxonomy governance.

Centralized or Decentralized? Loose or Tight?

When endeavoring to create robust taxonomy governance practices, the first decision to make is whether an organization needs a centralized or decentralized taxonomy governance structure. An organization needs to determine if they want to leverage a select group at the administrative level to determine modifications to the taxonomy (top-down), or allow end users to also suggest alterations (bottom-up).

Both models have their benefits and drawbacks, and these decisions are often contingent on the organization itself. If a taxonomy system will have several hundred end users who are all potentially voicing suggestions, having to filter through all these requests could quickly become unmanageable. Therefore, if an organization is confident in their ability to bring together a group of individuals who can represent the needs of all end users, then adhering to a centralized model makes sense.

Alternatively, if an organization is simply too diverse to bring together a truly representative group, or your end users expect the opportunity to provide feedback, the decentralized model would be more realistic.

Beyond deciding which model to adhere to, organizations must also consider how tightly or loosely they want to govern a taxonomy. This entails asking how strictly end users will be expected to adhere to the governance plan.

In a tight governance plan, there will be rigid controls on making changes to the taxonomy, and a clear chain of command through which authority has to be passed to make any changes. For organizations that are particularly risk averse, or who have more sensitive content, this may be the better solution. However, a rigid hierarchy can make it difficult for an organization to adapt and capitalize on changing business and content needs.

A loose governance plan enables individuals or lower level teams to have the autonomy and authority to experiment and make decisions about the taxonomy. An example of when this would be appropriate is when end users have the need to create individual tags for content.

While this plan allows for a taxonomy design to be more dynamic, having more end users make changes results in inconsistency within in the taxonomy design. The loose structure also heightens the possibility that a taxonomy design quickly becomes too large and disorganized to manage. At EK, we recommend that organizations start with a tight governance plan, recognizing that governance may shift to a looser structure as both the taxonomy and the role of end users change.

In any scenario, having an ill-fitting governance model that does not mesh with the organization can prove detrimental to the sustainment of the taxonomy design. Just like taxonomy design itself, one governance model does not fit all organizations.

Don’t Forget User Experience

Regardless of whether the user of the governance plan is a seasoned taxonomist or someone who has never heard the term before, a governance plan should always be designed for the non-technical user. In essence, the governance plan should be as easy to read and use to a brand new employee as it would be to those who created the taxonomy. There are many ways to augment the simplicity of a governance plan to make sure that anyone can use it to accomplish the tasks described within.

Most importantly, avoid overly technical jargon, and make sure that the document does not assume too much knowledge regarding key terminology or roles. Utilize graphics in order to convey essential points, such as governance procedures and roles (see example below). Doing so will cut down the amount of text needed to explain critical information.

Additionally, recognize that in most scenarios, a reader is not going to want (or have time) to read the entire document in order to figure out how to appropriately make changes to the taxonomy. Therefore, the governance plan should be designed as a “one-stop shop” reference guide, with sections linking to other parts of the document as necessary in order to guide readers through options and alternatives.

To further strengthen the usability of a governance plan, make sure that each section opens with the appropriate contextual information (e.g. definitions of what metadata, top-level, and lower-level values are, applicable roles). While this may result in a repetitive document, it alleviates the need for users to have to sift through other parts of the document to find the context they need so that they can move forward.

Take Away the Guesswork

A well thought-out governance plan ensures that important decisions regarding the taxonomy never have to be based on limited information or instinct. Rather, a robust governance plan provide users with as many tools as possible to ensure the process for making changes is calculated and logical. These tools can take a variety of forms, such as a simple list of questions or a formalized decision tree.

Metrics are another essential piece of eliminating guesswork within a governance plan. What might these metrics look like? It could be as simple as requiring that a proposed change reflect best practices (i.e. simple, clean language) or mandating that a change not conflict with other hierarchical relationships within the taxonomy. Other clients prefer more precise metrics such as search analytics or target numbers for content or users. The sample decision tree below provides some examples of the kinds of questions that can be used to guide clients through the decision making process.

Work in Tandem

A governance plan that is not applicable to the organization’s chosen technology is about as useful as no governance plan at all. The moment that the governance process is at odds with the technology in place is the moment that the future of a taxonomy is at risk. Therefore, selecting the right tools that align with a governance plan will provide the greatest business value with the least administrative burden.

Regardless of whether an organization has chosen a technology and subsequently written a governance plan or vice versa, the needs expressed in a governance plan need to align with the capabilities of the chosen tool. For example, a critical aspect of any governance plan will be creating, merging, and deleting terms, and managing relationships within the taxonomy hierarchy. Can the tool support the most common editing task, and can users change many terms at once? Another critical piece of a governance plan can involve having import and export functionality to pull values from lists or spreadsheets. In that case, it would be important to understand the robustness of the spreadsheet import and export functionality.

If the rules and tools don’t align, there may be opportunities to build a governance plan into the technology. At EK, we combine our taxonomy design efforts with our software development expertise in order to implement the workflows, roles, and other capabilities of a governance plan directly into an organization’s chosen tool. Building governance directly into your tool(s) ensures adherence to the plan and overall, makes it easier on the end user.

Ask for Feedback and Respond

Most importantly, seek out feedback on the governance model and plan. Although governance can support long-term ownership and responsibility, this can only be sustained by thoughtful and direct feedback from end users. Creating this culture of feedback doesn’t start with governance; rather, it should be present throughout a taxonomy design effort. When end users know that they are the primary focus in a taxonomy design, they have a sense of ownership and incentive to provide feedback because it is in their best interest to do so.

A simple tactic to create and sustain effective feedback loops is to document all decisions and archive all input in a common space that is accessible to end users. This demonstrates to end users that they are being heard, regardless of whether or not their suggestion has been implemented. As this repository gets built out, it can also serve as a reference point so that end users can see what changes have been made or not, and why.

Closing

In an ideal situation, a governance plan serves as a living document, one that is constantly leveraged to enable the evolution of an organization’s taxonomy design months, and even years into the future. Governance is the driving force that ensures that these changes are implemented in a logical and sustainable way. By utilizing governance to create these policies and procedures for change, organizations can be empowered to respond to feedback and bring the taxonomy to its best version possible.

Do you need help developing an effective taxonomy governance plan? Contact us at Enterprise Knowledge, we’d be happy to work with you.