Optimizing an online store’s category tree through tree testing

The challenge

Optimizing the category tree

At a store selling plants, decorations and garden accessories we identified that the category tree structure might be making products hard to find.

The store had a broad, deep assortment, which made the whole structure somewhat complicated. We had hypotheses about what could be improved, and to verify them we ran tree testing.

Context

A category tree shouldn’t “make you think”

“Don’t make me think” is the title of one of the most popular books on UX, and also one of the foundational principles of web design. Well over 20 years on from publication, the principle still holds. The point is that every website should be understandable, obvious and self-explanatory enough that it doesn’t “make visitors think” or work at it. 

One action that decidedly made this store’s users think was navigating the category tree and finding the subcategories and products they were after.

What is tree testing?

Tree testing lets you check how users navigate a category tree. Testers are given the tree of categories and subcategories, and their task is to locate specific items.

That shows whether users can find the information they need quickly and easily, and whether the information structure feels intuitive to them. Tree testing is usually run with a group of 30-100 people.

What we did

Running the tree testing

We had hypotheses about what might be problematic in the current category tree. To verify them we ran tree testing.

One of the tasks given to testers read: “Imagine you want to plant something new in your garden, but you need it to survive winter and low temperatures. Find a product in the store’s range that fits those requirements.”

We observed that only 69% of users completed that task and found the right subcategory.

Another task surprised us even more. Testers were asked to find “vases” in the category tree.

Only 34% of users managed it. “Vases” were buried so deep in the subcategory structure that reaching them decidedly “made you think.”

And here are a few quotes from testers:

• “Finding a vessel (vase) for cut flowers was a problem, because the menu only had pots and accessories. No other vessels or vases were listed anywhere, and there was no category for cut flowers at all.”
• “I had no idea where to look for a flower vase — to me it’s a decoration, but I’m not certain”
• “I wasn’t sure where I was supposed to look for a larger pot”
• “The vase took me longest. I clicked the right category first, but I didn’t expect it to be in a subcategory on top of that.”

Outcomes

A clear, intuitive category tree that’s easy to navigate

Building on the research findings, we reworked some category, subcategory and sub-subcategory names, changed their structure and contents, and added extra explanations in places.

We then ran another round of tree testing on the rebuilt tree. Users navigated the categories and found the products they were looking for markedly better.