How to Create a Card Sorting Template: Complete Guide

A card sorting template is a structured setup for your card sorting study. It includes the content items you want tested (usually 20-50 of them, written on digital or physical "cards"), a set of participant instructions, and a plan for collecting and analyzing results. When it's done well, a good template helps you understand how real users think about your content — so you can build navigation that actually makes sense to them.

What You'll Need

Before you start building your template, gather these basics:

Clear research questions — what specifically do you want to learn?
A content inventory — the features, pages, or topics you want participants to sort
15-30 target participants from your actual user base
A digital card sorting platform like CardSort, OptimalSort, or UserZoom
Spreadsheet software for planning your card set and analyzing results later

Step 1: Define Your Research Objectives and Scope

Your research objectives shape everything else in the template, so start here. What do you actually want to find out?

Try to narrow it down to 2-3 specific questions. Something like "How do users group our product features?" or "What categories make sense for our help center articles?" If you pile on too many objectives, participants get overwhelmed and your data gets muddy.

Your objectives also determine which sorting method you'll use (more on that in Step 3). Write down your core research questions before touching anything else — it keeps the whole process focused and prevents you from second-guessing decisions later.

Step 2: Create and Refine Your Card Set

This is where most of the work happens. You're aiming for 20-50 cards, and each one should represent a single, clear concept. No jargon. No overlapping ideas. Just plain language that your users would immediately understand.

Here's what trips people up:

Internal terminology sneaks in. You've been steeped in your product's vocabulary, so words that feel obvious to you might confuse participants. "CRM Integration" means nothing to most people — try "Connect with customer management tools" instead.
Cards are too vague or too specific. "Account Settings" and "Change Password" are at totally different levels of detail. Mixing them in the same study creates weird, unreliable groupings.
Descriptions run too long. Keep cards to a few words. Participants should be able to read and sort them quickly.

A good test: hand your card list to a colleague who isn't familiar with your project. If they can understand every card without asking questions, you're in good shape.

Step 3: Choose Your Sorting Method and Structure

You have three options, and the right one depends on your research goals.

Open card sorting is best for exploration. Participants create their own categories and name them however they want. This is great early in a project when you're trying to understand how users naturally think about your content. Plan for 15-20 participants to see reliable patterns emerge.

Closed card sorting tests an existing structure. You give participants a fixed set of categories and ask them to sort cards into those buckets. Use this when you're validating a navigation design or comparing alternatives. You'll want 20-30 participants here for more confident results.

Hybrid sorting splits the difference. You provide some starter categories but let participants create new ones if nothing fits. It's useful when you have a rough structure in mind but want to stay open to surprises.

Pick the method that matches your phase: exploring means open, validating means closed.

Step 4: Write Clear Participant Instructions

Your instructions need to explain the task without steering people toward any particular outcome. That's the tricky part — you want participants to sort based on their own mental models, not yours.

Good instructions cover:

What the study is about (in general terms)
How long it will take (usually 15-30 minutes)
Step-by-step directions for the sorting task
Whether they can create categories, rename existing ones, or leave cards unsorted

The tone matters. Make it feel low-pressure. Something like:

"Group these website features into categories that make sense to you. Create category names that describe each group. There are no right or wrong answers — we just want to understand how you'd naturally organize these items."

Avoid anything that hints at a "correct" answer or mirrors your company's internal structure.

Step 5: Set Up Your Template in a Digital Platform

Once your cards, method, and instructions are ready, it's time to put everything into your platform. Enter your card titles and descriptions, select your sorting type, and paste in your participant instructions.

A few settings worth paying attention to:

Randomize card order. If every participant sees cards in the same sequence, the first few cards tend to anchor how they think about the rest. Randomizing fixes this.
Add demographic questions. Collecting job role, experience level, or other background info lets you spot differences in sorting patterns across user segments.
Set category limits if needed. For closed sorts, make sure your categories are configured correctly before sending the study out.

Step 6: Test and Refine Your Template

Don't skip the pilot test. Run through the study with 2-3 people from your target audience before launching it for real.

Watch for:

Cards that consistently confuse people or get dumped into "other"
Instructions that raise questions or lead to unexpected behavior
Technical hiccups — broken links, display issues, anything that interrupts the flow
How long it actually takes (your estimate might be off)

Catching problems at this stage saves you from collecting a batch of unusable data and having to start over with new participants.

Step 7: Plan Your Data Collection and Analysis

Before you hit "launch," figure out your numbers and your analysis approach.

For quantitative patterns (clear groupings you can act on with confidence), aim for 15-30 participants. For qualitative insights (understanding why users think about things a certain way), 8-12 can be enough. If budget allows, recruit about 20% more than your target to account for incomplete responses and no-shows — they happen more often than you'd expect.

Decide ahead of time what you'll be looking at in the results:

Participant agreement rates (how often people put the same cards together)
Which categories are most and least popular
Card co-occurrence patterns (which cards consistently end up in the same group)

Having an analysis plan before data collection starts keeps you from fishing around in the results for meaning after the fact.

Pro Tips

Keep card text short. Two to seven words is the sweet spot. Longer descriptions slow participants down and add noise to your results.
Test on mobile. A significant number of participants will do your study on their phone. Make sure it works well on smaller screens.
Set honest time expectations. If your study takes 25 minutes, say so. Underestimating leads to frustrated participants who abandon halfway through.

Common Mistakes to Avoid

Too many cards. Going past 50 cards wears participants out. Completion rates drop, and the people who do finish tend to rush through the last batch. Keep it manageable.
Internal jargon everywhere. Acronyms and company-specific terms confuse outside participants and pull the results toward your team's mental model instead of your users'.
Mixed abstraction levels. If some cards are broad categories and others are tiny features, people won't know how to group them consistently.
Skipping the pilot. It's tempting to just launch, but one quick pilot round almost always surfaces something you missed.

Frequently Asked Questions

How long does it take to create a card sorting template?

Most of the upfront work takes a few hours — maybe half a day if you're being thorough. The bulk of that time goes into writing and refining your card set, which is worth getting right. Budget another hour or two for pilot testing. It feels slow, but catching problems early is far cheaper than restarting data collection.

What tools do I need to create a card sorting template?

At minimum, you need a card sorting platform and a spreadsheet. The platform (CardSort, OptimalSort, UserZoom, etc.) handles the participant-facing study and compiles results automatically. The spreadsheet is for planning your card set, tracking recruitment, and doing any custom analysis. Digital tools save you a huge amount of manual work compared to sorting physical index cards.

What are the most common card sorting template mistakes?

The big three: stuffing in too many cards, writing cards in language your users don't actually use, and mixing broad concepts with narrow ones in the same study. All three lead to messy, inconsistent data. A quick pilot test with a few real users will usually reveal these problems before they ruin your full study.

How do I know if my card sorting template is effective?

Look at a few signals. Are most cards landing in meaningful categories, or are a lot of them ending up in "miscellaneous" or "other"? Are participants finishing in a reasonable time without asking for clarification? Do similar cards consistently cluster together across different participants? When you see strong agreement and clean groupings, your template is working.

How many participants do I need for reliable card sorting results?

It depends on what you're after. If you want solid quantitative patterns you can confidently base design decisions on, aim for 15-30 participants — with 20 being a good minimum for statistical confidence. If you're earlier in the process and just want to understand how users think about your content, 8-12 participants can surface plenty of useful insights.

How to Create a Card Sorting Template: Complete Setup Guide