HomeCasa Card sorting results analysis to propose site navigation
Analyze open, closed or hybrid card sort data: standardize categories, measure participant agreement, find problem cards and propose an information architecture you can validate with a tree test. Designed for ChatGPT, Claude, Gemini and Microsoft Copilot. It includes 5 customizable variables and you can copy it for free in one click.
Advanced · Text · Pending editorial review
Prompt
Act as a UX researcher specializing in information architecture and card sort analysis.
Context:
- Site or product: {{site}}
- Card sort type: {{sort_type}}
- Number of participants: {{participants}}
- Study data (card, participant, assigned category): {{data}}
- Current navigation: {{current_navigation}}
Deliverables:
1. Data review: cards and participants included, incomplete or outlier responses and whether the sample size supports conclusions or only direction.
2. For open or hybrid sorts, a standardization table that merges equivalent category names created by participants, explaining each decision.
3. A card × category matrix with the number and percentage of participants who placed each card in each category.
4. Card pairs that most often appear together (similarity) and the natural groups that emerge.
5. Low-agreement cards: likely causes (ambiguous name, content that fits several places) and a recommendation (rename, cross-link from two sections, split).
6. Proposed architecture: top-level categories with labels, the cards in each one and a confidence level for every decision.
7. Comparison with the current navigation: what stays, what changes and the risk of each change.
8. A tree test plan to validate the proposal: 8-10 tasks, correct answer and success criterion.
9. Limitations of the analysis.
Constraints:
- Calculate only from the data provided and show how you reach each percentage; do not invent participants or results.
- If the sample is small, say so clearly and avoid strong conclusions.
- Recommend double-checking the calculations in a spreadsheet or card sorting tool.
- If the data is not in an analyzable format, ask me to restructure it before starting.
Format: tables for standardization, the matrix and the proposal; lists for problem cards and the tree test plan.How to use it
- Export your tool's results as card-participant-category rows. 2. Paste the data and your current navigation. 3. Review the low-agreement cards and run the proposed tree test before implementing.
Usage example
Site: city council website · Sort: open · 18 participants · 40 cards · Current navigation: City Hall, Services, Permits, News.
Sample output
The study rows haven't been pasted yet, so I won't calculate percentages; the tables below are ready to fill in as soon as I have them.
1. Data review
- Participants: 18. For an open sort, that's a useful sample for direction, not firm conclusions.
- Exclude responses with fewer than [FILL IN: minimum %] of cards sorted.
2. Category standardization (template)
| Participant labels | Standard category | Reason |
|---|---|---|
| [FILL IN: e.g., variants of "Permits"] | Permits and requests | Same concept, different name |
| [FILL IN] | News | [FILL IN] |
3. Card × category matrix
Each cell = participants who placed the card in that category ÷ 18. Example: 12 of 18 = 66.7%.
| Card | Permits | Services | City Hall | … |
|---|---|---|---|---|
| [FILL IN] | [n] ([%]) | [n] ([%]) | [n] ([%]) | … |
5. Low-agreement cards
If no category passes [FILL IN: threshold], review the card's name or cross-link it from two sections.
9. Limitations
With 18 people, a difference of one or two responses swings the percentages a lot. Double-check the math in a spreadsheet or your card sorting tool…
Key facts
- Category
- UI/UX design · Wireframes
- Output type
- Text
- Level
- Advanced
- Designed for
- ChatGPT, Claude, Gemini and Microsoft Copilot
- Verified in
- Pending editorial verification
- Variables
- Site or product, Card sort type, Number of participants, Study data and Current navigation
- Language
- English
- Price
- Free, no sign-up
- Updated
- Nov 7, 2026
Frequently asked questions
How many participants does a card sort need?
For stable patterns, at least 15-20 participants per user group is a common recommendation. With fewer, results are directional, and the prompt will flag that in the limitations.
What is the difference between card sorting and tree testing?
Card sorting reveals how people group content; tree testing checks whether they can find things in a specific structure. Teams usually use the first to propose and the second to validate.
What is this prompt for?
Analyze open, closed or hybrid card sort data: standardize categories, measure participant agreement, find problem cards and propose an information architecture you can validate with a tree test.
Which AI tools does it work with?
It is designed for ChatGPT, Claude, Gemini and Microsoft Copilot. We have not verified it in these tools yet, so results may vary between them.
How do I customize the prompt?
Click “Customize” and fill in the variables (Site or product, Card sort type, Number of participants, Study data and Current navigation). The text updates instantly and you can then copy it in one click. Values are stored only in your browser.
Do I need experience to use it?
It is advanced: it is meant for people with experience in the topic who can evaluate and adjust the result.
Is this prompt free?
Yes. Every HomeCasa Prompts prompt is free, with no sign-up or subscription. You can also use them in professional projects.
Can I trust the AI output?
Treat it as a high-quality draft: always check data, figures and important claims before publishing or making decisions.