Card Sorting Sample Size: How Many Participants Do You Need?

The most common question we hear: "How many participants should I recruit for my card sort?" Here's the science-based answer.

Quick Answer

Minimum viable: 15-20 participants Recommended: 30-50 participants Ideal: 50+ participants

But it depends on your study type and goals. Keep reading for specifics.

Sample Size by Study Type

Open Card Sorting

Recommended: 20-30 participants minimum

• More variability in responses
• Need larger sample for patterns to emerge
• Diminishing returns after 30-40 participants

Closed Card Sorting

Recommended: 30-50 participants

• Less variability (fixed categories)
• Can achieve statistical significance with more participants
• Stronger conclusions with 40+

Hybrid Card Sorting

Recommended: 25-40 participants

• Combines challenges of both approaches
• More participants = clearer trends

Sample Size by Research Goal

Exploratory Research

Goal: Discover how users think Sample Size: 15-25 participants

• Looking for general patterns
• Not testing specific hypotheses
• Saturation matters more than statistics

Validation Research

Goal: Prove/disprove a hypothesis Sample Size: 30-50+ participants

• Need statistical confidence
• Comparing alternatives
• Stakeholder buy-in requires numbers

Comparative Studies

Goal: Compare two structures Sample Size: 50+ participants (25+ per condition)

• Split between conditions
• Need statistical power
• Detect meaningful differences

Diminishing Returns

Research shows:

• 15 participants: Identify 80% of grouping patterns
• 30 participants: Identify 90-95% of patterns
• 50+ participants: Marginal improvements in pattern discovery

Exception: If analyzing by user segment, need 30+ per segment.

User Diversity Considerations

Homogeneous User Base

(e.g., Internal tools for accountants) Sample Size: 15-20 participants

• Less variability
• Smaller sample sufficient

Heterogeneous User Base

(e.g., Public-facing e-commerce) Sample Size: 30-50+ participants

• More variability
• Need larger sample
• Consider segmentation

Multiple User Personas

Sample Size: 15-20 per persona

• Test each persona separately
• Compare results across personas
• Total sample may be 45-60+

Budget vs. Sample Size

Tight Budget

15-20 participants: Still valuable

• Focus on qualitative insights
• Don't over-claim statistical significance
• Use for directional guidance

Moderate Budget

30-40 participants: Sweet spot

• Good statistical confidence
• Reasonable cost
• Reliable patterns

Enterprise Budget

50-100+ participants: Maximum confidence

• Segment analysis possible
• Strong stakeholder buy-in
• Publication-worthy rigor

How to Reduce Sample Size Needs

1. Pre-filter participants: Recruit only target users
2. Run pilot studies: Refine before scaling
3. Use closed card sorts: Less variability
4. Combine with other methods: Card sort + interviews
5. Segment analysis: 20 participants across 2-3 segments

Common Mistakes

Too Few Participants

❌ 5-10 participants

• Patterns may be artifacts
• Individual biases amplified
• Hard to draw conclusions

Too Many Participants

❌ 100+ participants (unless segmenting)

• Diminishing returns
• Wasted budget
• Analysis paralysis

Recruitment Cost Estimates

Based on $10 incentive per participant:

Our Recommendation

Start with 30 participants as your default:

• Balances cost and confidence
• Reliable patterns emerge
• Statistical analysis viable
• Segmentation possible (2-3 groups)

Scale up to 50+ if:

• High-stakes decisions
• Comparing alternatives
• Multiple user segments
• Budget allows

Scale down to 15-20 if:

• Exploratory only
• Tight budget
• Homogeneous users
• Combined with qualitative methods

Getting Started

Free Card Sort supports unlimited participants on the free plan. Start with 30 participants and scale up if needed - no additional cost!

Run your card sort free at freecardsort.com