Skip to main content
Irrevo Best Practices

Creating a taxonomy

A taxonomy is classification system. It is a way to categorize your knowledge base content and make it easy for your users to find what they need. Without a classification system, valuable knowledge gets lost and can be difficult to access.

Get to know your users

Before diving into categories and labels, you need collect information about your users and how they think.

  1. Who are your users?
    Is your knowledge base geared toward customers, customer service representatives, or operations personnel? Maybe you plan to use your knowledge base for multiple users. Clearly define who your users are. If there are multiple users, clearly define the user groups.
  2. Based on the user group(s), what is the purpose of your knowledge base?
    Is the content intended for training, general information, or troubleshooting? The way you categorize your content will be largely based on your intention for the users.
  3. How do your users search for knowledge?
    Research the way your current users navigate the content to look for the information they need. You can also run simple test studies to determine how your users ask questions and search for information.

Organize your content

Once you really get to know your knowledge base and your users, organize your content into broad categories. For example, you may find a lot of content pertains to “billing,” “ordering,” “troubleshooting,” or “account support.”

After sifting through all your content, you may decide that some categories can be combined. For example, you may decide that “billing” and “ordering” are very similar and can be combined into one category, “orders and billing.”

If you’re having trouble telling the different between “billing” and “ordering,” try a card sort exercise with your agents. Card sorting is a user-focused approach to organizing information.

Decide on a level of depth

Based on the quantity and complexity of each category, decide on a level of depth.

For example, your “billing” category may include a lot of related content; so, it might make sense to create a second level of subcategories such as “online bill pay,” “refunds,” and “statements.”

If you have a very large knowledge base,  it may necessary to create a third level for some of your subcategories.

For example, if the refund process for individual customers is different than that for businesses, it may be helpful to divide “refunds” further into “individuals” and “businesses.”

A taxonomy is classification system. It is a way to categorize your knowledge base content and make it easy for your users to find what they need. Without a classification system, valuable knowledge gets lost and can be difficult to access. In fact, a recent study showed that only 53% of organizations have effective search capability in their knowledge base.

A taxonomy is not an arbitrary, however. It takes a bit of thought and planning to make a meaningful system to accurately represent your knowledge base.

In this article, we discuss how to create an effective taxonomy for your knowledge base.

Get to know your users

Before diving into categories and labels, you need collect information about your users and how they think.

  1. Who are your users?
    Is your knowledge base geared toward customers, customer service representatives, or operations personnel? Maybe you plan to use your knowledge base for multiple users. Clearly define who your users are. If there are multiple users, clearly define the user groups.
  2. Based on the user group(s), what is the purpose of your knowledge base?
    Is the content intended for training, general information, or troubleshooting? The way you categorize your content will be largely based on your intention for the users.
  3. How do your users search for knowledge?
    Research the way your current users navigate the content to look for the information they need. You can also run simple test studies to determine how your users ask questions and search for information.

Organize your content

Once you really get to know your knowledge base and your users, organize your content into broad categories. For example, you may find a lot of content pertains to “billing,” “ordering,” “troubleshooting,” or “account support.”

After sifting through all your content, you may decide that some categories can be combined. For example, you may decide that “billing” and “ordering” are very similar and can be combined into one category, “orders and billing.”

If you’re having trouble telling the different between “billing” and “ordering,” try a card sort exercise with your agents. Card sorting is a user-focused approach to organizing information.

Decide on a level of depth

Based on the quantity and complexity of each category, decide on a level of depth.

For example, your “billing” category may include a lot of related content; so, it might make sense to create a second level of subcategories such as “online bill pay,” “refunds,” and “statements.”

If you have a very large knowledge base,  it may necessary to create a third level for some of your subcategories.

For example, if the refund process for individual customers is different than that for businesses, it may be helpful to divide “refunds” further into “individuals” and “businesses.”

Test your taxonomy

Instead of wondering whether or not your taxonomy will be effective once it’s live, why not test it with focus groups first? Focus groups can help sort out any major issues and give you more insight as to how users navigate the content. Once the content is live, be sure to continue testing in order to adapt to changes in user patterns and needs.

Taxonomies gone bad

Be on the lookout for these potential blunders:

  1. Too many categories
    When information is spread too thin, it can get lost and users may not be able to find what they need. For example, “billing” and “payments” may be too similar to separate. Don’t make your users guess where they need to hunt.
  2. Too few categories
    When information is too concentrated, it can get lost and users may be overloaded. For example, if a user needs to log into his/her account to update a password or address, it’s not useful to lump all “my account” information under “order status” or “billing.”
  3. Too many or too few levels of depth
    Smaller knowledge bases may not need subcategories, but large knowledge bases might. You may be tempted to create as few levels of depth as possible to avoid having your users spend more time clicking through your content. However, studies show that users don’t mind clicking through information as many times as necessary, provided that the navigational path makes sense. UX Myths busted the myth that pages should be accessible in three clicks by stating that “…the number of necessary clicks affects neither user satisfaction, nor success rate. That’s right; fewer clicks don’t make users happier and aren’t necessarily perceived as faster.”Don’t forget that you can always interlink information with each article. As long as you provide a logical pathway for the user to find needed information, level of depth will not threaten the ease of access.

Get help from the pros

Take a look at some support websites for companies of your size and see what they do. Pay attention to how they structure their content, what you like about it, what you don’t like about it, and what you can learn from it.