How should we represent content that appears in more than one place in our tree?

For example, suppose we have a Frequently Asked Questions (FAQ) page for a product, and that FAQ page is available from both the Products section and the Support section, like this:

• Products
  • Wizzo-matic
    • Features
    • FAQ
• Support
  • Downloads
  • FAQs
    • Wizzo-matic FAQ
      
      

It could be that this FAQ page is actually duplicated – each section has a separate page with the same content. Or (and this seems to be more common), there is only one actual FAQ page (say, the one in the Support section), and the one in the Products section is just a link to it.

In the tree we’re going to test, do weinclude the FAQ topic in both sections, or do we remove the one in Products because it’s not a “real” content page?

In general, we recommend including both entries in the tree, because even if the Product-FAQ link just links to the Support section, it’s still a choice that we give the user when they go to the Wizzo-matic product page. If wedidn’t include the Product-FAQ topic in our tree, we would be withholding information from the test participant that a visitor to the real site would have.

Later, when we analyze the results, we will definitely want to see if participants took the Products-FAQ path or the Support-FAQ path. If we see that 90% of them go first to the Products section for the FAQ, that suggests that the FAQ should move to live there, and the Support-FAQ topic should just be a link to it.

If we choose not to include both entries in our tree, we can still check the results to see which path the participants took. If a sizable fraction of them went to the section without the FAQ topic, that suggests that we should at least include a cross-link there when we design the real site.

What if the duplicated content is not just a single topic (like the FAQ page above), but an entire subtree of topics? Should we duplicate the entire subtree for testing?

For example, suppose we have a My Account section in our site, which lets users create an account, log into that account, and see a my-account FAQ. It “lives” in the Support section, but we’ve also added a shortcut in the main navigation bar (a utility link and pop-up menu in the upper right of the header):

Here’s how we might represent this in a text tree:

• Products
• Support
  • My account
    • Sign up
    • Login
    • FAQ
  • Downloads
  • Forums
• About us
• My account
  • Sign up
  • Login
  • FAQ
    
    

Notice that My Account appears twice in the tree, because it appears in two places in the global navigation.

Again, in general, we recommend including both subtrees in the test, because that’s what the user would see in the real site.

The only time we may not want to duplicate subtrees is if the subtree is very large, and it becomes tedious to keep both instances in sync as we develop the tree. In that case, we may want to prune the duplicate that we think is going to get less traffic during the test.