Card SortTree Test

Which should you do first? Card sort or tree test?

4 minute read time

You know you want to improve your website navigation but you’re not sure where to start.

Should you start with the card sort or the tree test?

For most situations you want to start with a tree test and then an open card sort followed by another tree test.

After that it kind of depends on what results you found in the first three tests.

Let’s walk through why those three tests first and what do afterwards.

Why a tree test first?

Let’s answer that question with another question.

What if your navigation isn’t actually broken like you think it is?

Are you absolutely positive that it is an issue?

It might not be.

You never know.

Crazier things have happened.

The reason running a tree test is recommended as the first step is because it’s going to give you some answers to these questions.

It’ll give you some data points showing you if your audience is getting lost and if so, where.

Wouldn’t it be nice to know what’s working in your navigation and what isn’t?

That’s what a tree test is going to tell you.

Also, you can use the data you gather from this first test as a benchmark to measure against any changes. This will allow you to make sure you actually improve your website navigation for your audience. And it will give you data on how much you improved it!

Why an open card sort next?

Once you validate the need to improve your navigation and you spot the problem areas, you’re ready to begin fixing the problems.

That’s where the open card sort comes into play.

You’re going to let your audience show you how your navigation should be organized and what labels you should use to use to describe each category.

Your audience will use their own words. They’ll also open up their mental model of the way they think about your content.

If you’re a furniture company, do they think about your products in terms of which room it would go in? Or do they think about in terms of which type of wood? Or maybe they think about it in terms of type of furniture (couches, beds, etc.).

However they think about it, they’re going to show you in an open card sort.

You’re going to get valuable information, not only about how to organize your content, but how your audience thinks about it.

On top of all this information, you’re going to get information about what words they would use to describe their organization of your content.

Going back to the furniture company, let’s say you discovered most of your audience thought of your products in terms of rooms. You’ll get data back about what they call each room.

With this information you can use labels and groupings that makes sense to them.

Why another tree test?

You can take all those data points from the open card sort and come up with a revised navigation that you think will be best.

Now, it’s time to see if you improved things for your audience.

Running another tree test will give you an idea of how your audience performs completing the same tasks as the first tree test.

Were they more successful?

Did you uncover any other problem areas?

Depending on the results you get from this second tree test your user research could take some different paths.

Let’s explore what those could be.

The second/third tree test was a complete success!

Let’s say the your tree test showed a complete success - your audience was super successful with no problem areas showing up.

If so, implement that new navigation and website structure. You’re onto other user experience improvements. You nailed this one.

The second/third tree test was partially successful.

Part of your navigation outperformed the other part but you also found some problem areas.

No big deal.

You’re half way there!

Do you have an idea, based on the data from the original card sort what might be the solution?

If so, try that out with another tree test. You may be onto something.

If you’re not sure what went wrong, you’ll want to run another card sort. The type of card sort you choose depends on where exactly in the tree your audience had problems.

Was it on the main navigation items? Or was it deeper in the tree?

If it was on the main navigation items you’ll want to run a hybrid study.

Create the categories that performed well the first time. Do not include the navigation categories that were unsuccessful.

In this test you’re going to allow your audience to group the remaining content in a way that they think is best. This will give you ideas on better navigation naming options and organization.

If the issues were deeper in the tree you’ll want to run a closed study.

Keep what was working well and dive into the underlying organization to narrow in on a better solution.

Whether it’s a closed or hybrid study, this test should be faster and easier for them, with the focus being on the problem area.

The results from this study will give you some more insights that will hopefully lead you to a successful categorization and labeling direction.

Use the results to come up with revised navigation and labels and then test it with a tree test.

The second/third tree test was a failure.

If you did worse on this tree test than the first one, go back to your card sort results and see if you can review the results and come up with a better solution.

You can then run a tree test checking this solution.

If the results don’t help you, it might be an issue of scattered and unhelpful results. In that case you should re-do the card sort. Take a look at these quick tips to ensure your card titles produce good results for your card sort.

You can’t go wrong.

Whether your first attempt at an improved navigation was successful or not, you win either way.

Each time you learn more about your audience, the way they think about your content, and the words they use.

This is all super valuable and will help you with not only your navigation but in serving them in all areas of your business.

If you’re not sure whether your navigation is working well or not, run a quick tree test and get some answers.

There is no way to lose here.

You’ll get answers.

You’ll come to understand your audience better.

And you’ll have some data to back up your argument as to whether you should change your navigation or not.

This is a win-win situation.