To silo or not to silo: How to tell if you need a Corporate Taxonomy

Many data managers and those who rely on their digital work struggle with defining the taxonomy of the data under their control. Taxonomy is the science of classifying information into groups or classes that share similar characteristics and is required for meaningful information management and is critical to effective findability. The overall scheme for organizing information is to solve a business problem based on user needs and to show correlations between subjects.


For many, there is a tendency to want a universal solution: a central corporate taxonomy that drives consistency and ease of use. Others see value in maintaining separate schema for various datasets and use cases. Then there is a hybrid of the two: core of universal terms across data sets but ultimately customized taxonomy for each situation.


While there are certain circumstances under which each of these solutions is ideal, there are key elements to consider when selecting a solution.



Flexibility is Key

There is no “one size fits all” solution for data taxonomy. Finding the ideal fit for an enterprise is a time-consuming challenge. However, doing the work upfront to establish a method appropriate to the scope of the issue will more than pay off in time saved further down your timeline. Front-loading your process is invaluable.


Discuss use cases with those who will need access to the data. Find commonality and consistency of language and approach and use that to fuel decisions about how centralized your schema should be, what exceptions might exist, and how to describe and cross-reference those exceptions.



The Data is Not Yours

It’s easy to feel proprietary about the data under your command and the carefully crafted taxonomy that was developed. Remember that the ultimate goal is usefulness. Your taxonomy is a service to those who need the data: it makes it not just possible, but easier for users to make use of the assets they need.


Think of the users and those managing the data as your customers and greatest allies. Listen to their needs, discover their pain points, and work with them to find ways to ease the flow for them. Relieving their pain points will ultimately relieve the data problems further down the chain.



Tools Shape the Solution

Unfortunately, the perfect software doesn’t exist. Since customizing a solution to every need of your users is unrealistic, the best bet is to find a Venn overlap between the best of what you currently have, clarity of process to all users, and serving the needs of the owners of the data. Then use this as your basis to derive a solution map to find your tool.


Different tools will have varying benefits and drawbacks, and these too can influence the architecture of your taxonomy. Work with those who know the tools best and with those who know the data best to develop means of access and inform your decision making.



Putting it All Together

Combining tools with flexibility and ownership means that you may have different solutions for different users, even if the taxonomy is primarily centralized and specific. Use informs taxonomy, and it won’t matter how elegant your structure is if it remains opaque or confusing to users. It is impossible to anticipate every possible usage case or information need, so stay calm and remember that taxonomy is a living document that learns and grows with the needs of its users.


A central, corporate taxonomy meets many needs, and will be a success if the time is taken to ensure that factors around user base, use cases, accessible taxonomy language, and finding the right tool for your content are carefully considered and enacted upfront. Resist the urge to move at lightning speed, as all the time taken at the front end to determine the correct solution for your data will speed up user adoption, findability, and access.

Asset 7
  • Washington, D.C.
  • Berlin
  • Los Angeles
  • Minneapolis
  • New England
  • New York