Want to build a great site map? Look outside your team.
Open card sorting can reveal the insights that make your website a joy to use … or easy to hate.
At work recently, I was in a meeting to talk about a subsection of the company’s website. Everyone seemed to have a different, conflicting opinion about how to reorganize it — that’s when I raised the question: how can we sort out this content if we’re not even its audience?
As practical as it might seem to ask the question, it’s surprising how many projects move forward without even considering it. But there’s no need to guess if the answers lie with your users. Understanding how they associate information and group things can go a long way to informing your navigation taxonomy and sitemap decisions.
In this article, I’ll walk you through an easy exercise I did for an independent project — an insurance website targeted at millennials — that helped me create a sitemap based on user feedback, not my own guesswork. The method, open card sorting, is a simple UX research technique where participants are asked to group information and create names for those categories. It’s a powerful way to approach design for information architecture, workflows, menu structures, and site navigation paths.
Step 1: Find users
Going through my network of friends and LinkedIn contacts, I identified seven individuals who fit my general user profile. The only information I provided was that they’d be doing an exercise that would involve organizing products based on what made sense to them. There was no right or wrong answer.
Step 2: Set up the test
Using an online card sorting tool, OptimalSort, I created 20 product types that I wanted my test participants to organize.
I asked each person to group the cards according to what made the most sense to them. They were then asked to provide names for these groups.
Step 3: Look for patterns
Once the sessions were done, I looked at the results to identify patterns. Tools like OptimalSort can help you quickly identify these patterns but you can also just do this on your own. For example, most of my users created a category that was a variation of “Life” or “Life Insurance” and it contained products that were mostly the same (at least 80%), with only a few outliers. What does this tell me? “Life” makes sense to them as a top-level category containing these products.
The similarity matrix in OptimalSort also let me know what proportion of my participants grouped any 2 cards in the same category. For each pair of cards, the intersecting cell shows the percentage of participants who grouped the cards together. The darker clusters are great places to identify potential groupings.
Step 4: Identify problem spots for follow up
Some products were categorized inconsistently, which meant they weren’t clear to users. For example, insurance related to vehicles was inconsistently lumped with travel insurance. Was insurance about protecting an object owned, or an experience purchased? I believe questions like this warranted further research.
Another problem that became clear: certain product names weren’t obvious. For example, some users had a difficult time creating category names when they didn’t know what “business interruption” or “agricultural” insurance meant.
Step 5: Create your taxonomy
Based on the patterns I identified and working through the problematic product names, I now have a taxonomy I have a much higher confidence level in, compared to, say, making guesses on my own as to which products belong with one another.
While it’s tempting to blast through something like a sitemap, consider how important it will be to your design, and how much influence it will have.
Taking the time to understand your user and conducting a simple open card sorting exercise like this can make the difference between launching a site that is easy to use vs. one that will frustrate and confuse.
