15 Real Card Sorting Examples from Successful UX Projects

Card sorting gives you raw data about how people group things together. But raw data alone doesn't tell you what to do next. You need to see what the patterns actually look like in practice — what worked, what surprised people, and how the results translated into better navigation.

That's what this article is for. We've pulled together real examples of card sorting studies across e-commerce, banking, SaaS, intranets, and more. The first five are detailed walkthroughs with full study setups and results. After that, we cover ten more patterns worth knowing about, pulled from studies across education, healthcare, government, and other industries.

A few things you'll see come up again and again: users organize by what they're trying to do, not by how your company is structured. They gravitate toward plain language over jargon. And they almost always settle on 4-7 main categories — not the 15-20 that many teams start with.

Quick Example Overview

Example 1: E-commerce Site Navigation

An online clothing retailer had 35 product categories crammed into their navigation menu. Bounce rates were high, and shoppers kept telling support they couldn't find what they were looking for.

Study Setup

Type: Open card sort Cards: 35 product categories

- Men's T-Shirts
- Women's Dresses
- Kids' Shoes
- Athletic Wear
- Formal Wear
- Casual Shirts
- Winter Jackets
- Summer Dresses
- Business Attire
- ...and 26 more

Participants: 30 shoppers (mixed demographics) Instructions: "Organize these products into groups that make sense to you when shopping. Name each group."

Results

Original Structure (confusing):

- Men's Clothing (22 subcategories)
- Women's Clothing (25 subcategories)
- Kids' Clothing (15 subcategories)
- Accessories (12 subcategories)

New Structure (from card sort):

Main Navigation:
├─ Shop by Occasion
│  ├─ Casual & Everyday
│  ├─ Work & Professional
│  ├─ Athletic & Outdoor
│  └─ Formal & Special Events
├─ Shop by Person
│  ├─ Women
│  ├─ Men
│  └─ Kids
└─ Sale & New Arrivals

Outcome

The new navigation led to a significant increase in product page views, a noticeable drop in bounce rate, and higher conversions. Shoppers told the team the navigation "made sense now" — which is exactly the kind of feedback you want but rarely get about navigation.

Key Insight

Shoppers wanted to browse by occasion (casual, work, athletic) rather than just gender. This wasn't obvious to the internal team, but it came through clearly in the card sort. People shop with a purpose in mind, and the navigation should reflect that.

Example 2: Mobile Banking App

A banking app had 28 features spread across 6 tabs. Users kept telling support they couldn't find basic things like "Pay a Bill" or "View Statements." The team suspected the navigation was organized around banking concepts instead of user goals, so they ran a hybrid card sort to test that theory.

Study Setup

Type: Hybrid card sort Cards: 28 banking features

- Check Balance
- Transfer Money
- Pay Bills
- Deposit Check
- View Statements
- ATM Locator
- Budget Tracker
- Savings Goals
- Investment Portfolio
- Customer Support
- Security Settings
- Card Controls
- ...and 16 more

Suggested Categories (for hybrid sort):

- Accounts
- Payments
- Tools
- Settings

Participants: 25 active banking app users Instructions: "Organize these features into the provided categories, or create new categories if needed."

Results

Most participants rejected or renamed the suggested categories. Here's what they came up with instead:

├─ My Money (38% created this)
│  ├─ Check Balance
│  ├─ View Statements
│  └─ Account Details
├─ Move Money (42% created this)
│  ├─ Transfer Money
│  ├─ Pay Bills
│  └─ Deposit Check
├─ Plan Ahead (35% created this)
│  ├─ Budget Tracker
│  ├─ Savings Goals
│  └─ Bill Reminders
└─ Settings & Help (45% created this)
   ├─ Security Settings
   ├─ Customer Support
   └─ Card Controls

The app went from 6 tabs down to 4 main sections based on these results.

Outcome

Task completion went up dramatically — roughly from the mid-60s to nearly 90%. Time to complete common tasks dropped by about a third. App store ratings climbed noticeably, and support calls went down.

Key Insight

Users think in terms of actions ("Move Money") rather than banking terminology ("Transactions"). Nobody in the card sort created a category called "Transaction Module." They said things like "Move Money" and "Plan Ahead." That's the language your app should use too.

Example 3: SaaS Product Dashboard

A project management tool had 42 features buried in nested menus. New users weren't discovering key functionality, which was killing activation rates. The features were organized by the engineering team's internal structure — not by how people actually work.

Study Setup

Type: Open card sort Cards: 42 product features

- Create Project
- Task Board
- Gantt Chart
- Time Tracking
- Team Chat
- File Sharing
- Calendar View
- Reports Dashboard
- Notifications
- User Permissions
- Integrations
- API Access
- ...and 30 more

Participants: 20 product managers and team leads (target users) Instructions: "Imagine you're using this tool for the first time. How would you group these features?"

Results

Top User-Created Categories:

1. Project Work (85% agreement)

   • Create Project, Task Board, Gantt Chart, Calendar View

2. Team Collaboration (78% agreement)

   • Team Chat, File Sharing, Comments, @Mentions

3. Tracking & Reporting (72% agreement)

   • Time Tracking, Reports Dashboard, Progress Charts

4. Settings & Admin (88% agreement)

   • User Permissions, Integrations, API Access, Billing

Implemented Structure

Sidebar Navigation:
├─ Projects (main work area)
├─ Team (collaboration features)
├─ Insights (reporting & analytics)
├─ Apps (integrations)
└─ Settings (admin functions)

Outcome

New user activation nearly doubled. Feature discovery improved across the board, onboarding time dropped significantly, and customer satisfaction scores went up. The biggest win was that new users were actually finding and using features that had been hidden in submenus before.

Key Insight

Users wanted a clean, focused workspace with advanced features tucked away but still accessible. The card sort revealed a natural split between core work features and secondary/admin features that the team hadn't seen from the inside.

Example 4: Corporate Intranet

A company with 10,000 employees had an intranet with 80+ pages organized in a massive alphabetical list. Employees constantly complained they couldn't find basic things like benefits info or IT support.

Study Setup

Type: Open card sort Cards: 50 most-visited pages

- Submit Time Off
- Health Benefits
- 401(k) Information
- Company News
- IT Help Desk
- Office Locations
- Employee Directory
- Training Courses
- Expense Reports
- Payroll Information
- Company Policies
- Department Contacts
- ...and 38 more

Participants: 40 employees (various departments and tenure levels) Instructions: "Organize these intranet pages into groups that would help you find what you need quickly."

Results

Original Structure: 80+ pages in an A-Z list (not useful for anyone)

New Structure (from card sort):

Quick Links (Dashboard):
├─ For Me
│  ├─ My Benefits
│  ├─ My Time & Pay
│  └─ My Career
├─ Need Help
│  ├─ IT Support
│  ├─ HR Questions
│  └─ Facilities
├─ Stay Informed
│  ├─ Company News
│  ├─ Events Calendar
│  └─ Announcements
└─ Resources
   ├─ Policies & Forms
   ├─ Training
   └─ Employee Directory

Outcome

Task success rates roughly doubled. Average time to find something dropped from minutes to under a minute. IT tickets about "can't find X" fell sharply. Employee satisfaction with the intranet went from poor to good.

Key Insight

Employees wanted task-based organization ("Submit time off") rather than departmental organization ("HR > Time Off > Submit Request"). Nobody thinks in org chart terms when they just need to file an expense report.

Example 5: Help Center Redesign

A SaaS company's help center had 150+ articles, but customers still contacted support for basic questions. The existing categories were organized by product features, which made sense internally but not to customers trying to solve a problem.

Study Setup

Type: Open card sort Cards: 30 most-searched help articles

- How to Reset Password
- Billing & Payments FAQ
- How to Export Data
- Account Security Setup
- Team Member Permissions
- Integration Setup Guide
- Troubleshooting Errors
- Mobile App Guide
- API Documentation
- Feature Tutorials
- ...and 20 more

Participants: 25 customers (mix of new and experienced users) Instructions: "You need help with our product. Organize these topics into groups that would help you find answers quickly."

Results

User-Created Categories (with agreement %):

1. Getting Started (82% grouped these together)

   • Account Setup, First Steps, Basic Features

2. Common Questions (75% agreement)

   • Password Reset, Billing FAQ, Account Settings

3. Advanced Features (71% agreement)

   • API Docs, Integrations, Custom Settings

4. Troubleshooting (88% agreement)

   • Error Messages, Common Issues, Bug Reports

5. Mobile & Apps (66% agreement)

   • Mobile Guide, Desktop App, Browser Extensions

Implemented Structure

The team added search tags and reorganized around these user-created categories:

Help Center:
├─ Getting Started (for new users)
├─ Common Questions (FAQ-style)
├─ Features & How-To (tutorials)
├─ Troubleshooting (problem-solving)
└─ Developers (API docs)

Outcome

Support ticket volume dropped noticeably. Help article views went up — people were actually finding and reading the articles now instead of just emailing support. Customer satisfaction improved, and average resolution time came down.

Key Insight

Users group help content by their goal (troubleshooting, learning, reference) rather than by product features. When someone has a problem, they're not thinking "which feature is broken?" — they're thinking "something isn't working and I need to fix it."

More Card Sorting Patterns

The five examples above give you detailed study setups to work from. But card sorting shows up across a lot more contexts. Here are ten more patterns worth knowing about, each with a useful takeaway you can apply to your own work.

Educational platforms — An online learning platform with 200+ courses found that learners prefer subject-based categories (Math, Science) over skill-level groupings (Beginner, Advanced). People browse by topic first, then filter by difficulty. If you're organizing educational content, lead with subject matter.

Recipe websites — A food blog with 500+ recipes discovered that home cooks organize by meal type (Breakfast, Dinner) and dietary needs (Vegetarian, Gluten-Free), not by cuisine. This makes sense when you think about it — most people are answering "what should I make for dinner?" not "I'd like something Thai today."

Fitness apps — A workout app with 60 exercises found users preferred grouping by body area (Upper Body, Core) over equipment type. People think about what they want to work on, not what machines are available.

Travel booking sites — A travel platform with 30 booking features found users wanted a trip timeline structure (Before Trip, During Trip, After Trip) instead of categories organized by service type (Flights, Hotels, Car Rentals). The trip itself is the mental frame, not the vendor.

News websites — A local news site found that some articles genuinely belong in multiple categories. Their card sort showed low agreement on where certain stories should live, which led them to implement a tagging system alongside traditional sections. If your card sort shows disagreement, that's not a failure — it's telling you something about your content.

Government portals — A city government website found that residents organize services by life events (Moving to the city, Having a baby, Retiring) rather than by department (Public Works, Health Department). Citizens don't know or care which department handles what.

Design resource libraries — A design agency with 200+ resources found that designers wanted project phase categories (Research, Ideation, Production) rather than file-type organization (Templates, Icons, Photos). Workflow trumps format.

Medical patient portals — A hospital portal found that patients strongly preferred plain language ("Talk to My Doctor") over medical terminology ("Secure Messaging"). This was one of the clearest results in any of these studies — patients don't speak healthcare jargon, and they shouldn't have to.

Real estate websites — A property listing site found that users created priority levels (Must-Have, Nice-to-Have) rather than feature categories when organizing search filters. Buyers think in terms of deal-breakers and bonuses, not property attributes.

Podcast apps — A podcast discovery app found high disagreement across genre categories, which told them that rigid genre buckets don't work for podcasts. They moved to a multi-tagging system instead. When users can't agree on where something belongs, the answer might be "it belongs in more than one place."

Proven Patterns Across These Studies

People Organize by What They Want to Do

This came up in almost every example. Users group things by their goals, not by your product's technical structure or your company's org chart.

• Banking app: "Move Money" instead of "Transactions"
• Intranet: "Submit Time Off" instead of "HR Forms"
• Help Center: "Getting Started" instead of "Features List"

If your current navigation is organized around internal concepts, a card sort will almost certainly tell you to flip it around.

Plain Language Beats Jargon Every Time

Medical portals, SaaS platforms, government sites — it doesn't matter the industry. Users pick everyday words over technical terms.

• Medical portal: "Talk to My Doctor" instead of "Secure Messaging"
• SaaS: "Team" instead of "Collaboration Suite"
• Government: "Having a Baby" instead of "Birth Registration Services"

4-7 Main Categories, Not 15-20

Users naturally settle on 4-7 main groupings. This lines up with what we know about working memory — people can hold about 5-9 chunks of information at once. If your navigation has 12+ top-level items, it's probably too many.

• Banking app: 6 tabs became 4 sections
• E-commerce: 8 mega-menu columns became 4
• Intranet: 12 sections became 4

When People Disagree, It Means Something

Low agreement in a card sort isn't a problem to solve — it's a finding. When participants can't agree on where something belongs, that content probably needs to live in multiple places. Tags, cross-links, and flexible navigation handle this better than forcing everything into one bucket.

Further Reading

• What is Card Sorting? Complete Guide
• Card Sorting (UX Glossary)
• Information Architecture (UX Glossary)
• How To Run Your First Card Sort Study

Frequently Asked Questions

How do you choose between open, closed, or hybrid card sorting? It depends on where you are in the process. If your current navigation is a mess and you need fresh ideas, run an open card sort — let participants create their own categories from scratch. If you've already got a structure you think might work, use a closed sort to test it. And if you want to propose some categories but still leave room for surprises, go with a hybrid sort. The banking app example above used a hybrid approach and it worked well because the team had a hypothesis but wanted to see if users would push back on it (they did).

What is the optimal number of cards and participants for card sorting? Somewhere around 30-50 cards and 20-30 participants tends to hit the sweet spot. The examples in this article ranged from 28 cards to 50 cards, with 20-40 participants, and they all produced clear patterns. Go below 20 cards and you won't see meaningful groupings emerge. Go above 60 and participants start getting fatigued, which muddies your data.

What should you do when card sorting results show high disagreement? Don't panic — disagreement is useful information. It usually means that certain content genuinely fits in more than one place. The podcast app example ran into exactly this: users couldn't agree on genre categories, so the team built a multi-tagging system instead of forcing rigid buckets. If you see agreement below 60% on certain items, consider cross-linking, tagging, or putting that content in multiple sections.

Can card sorting work for specialized industries with technical content? Absolutely. The examples here span healthcare, government, finance, and B2B software. The method works because it taps into how people naturally organize information in their heads — and that's universal. The key is recruiting actual users from your domain. Don't run a card sort for a medical portal with college students; recruit actual patients.

How long does it take to see results from card sorting implementations? Most teams see changes in their metrics within a few weeks of launching the new structure. The biggest shifts happen right away — when your navigation suddenly matches how people think, they find things faster on day one. The trailing indicators like support ticket volume and satisfaction scores take a bit longer to move, usually 4-8 weeks.