Compare

Compare enables side-by-side analysis of products, concepts, or changes over time. Use it for A/B testing, competitive analysis, and trend tracking.

Comparison Types

Product vs Product

Compare two different products:

• Preference distribution
• Attribute-by-attribute ratings
• Segment-specific preferences
• Switch likelihood

Before vs After

Track impact of changes:

• Perception shifts
• Preference changes
• Segment movement
• Overall improvement metrics

Concept Testing

Compare multiple concepts:

• Ranked preference
• Concept clarity
• Implementation priority
• Risk assessment

Running a Comparison

1. Go to Compare in the app
2. Select comparison type
3. Choose items to compare (products, sessions, concepts)
4. Configure parameters
5. Click Run Comparison

Configuration Options

| Option | Description | | ------------- | -------------------------------- | | Sample Size | Number of AI respondents | | Matched Panel | Use same personas for both sides | | Attributes | Specific dimensions to compare | | Segments | Focus on particular segments |

Understanding Results

Preference Matrix

Visual representation showing:

• Overall preference split
• Confidence intervals
• Segment breakdowns
• Statistical significance

Attribute Comparison

Side-by-side rating on dimensions like:

• Value for money
• Quality perception
• Trust
• Innovation
• Usability

Key Differences

AI-summarized insights about:

• Most significant differences
• Segment-specific preferences
• Recommendations

Use Cases

Product Iteration

1. Run Discovery on current version
2. Make changes
3. Run Compare (before vs after)
4. Measure improvement

Competitive Intelligence

1. Create competitor product profiles
2. Run Compare (your product vs competitor)
3. Identify strengths and gaps
4. Prioritize improvements

Feature Prioritization

1. Define feature concepts
2. Run Compare across all concepts
3. Rank by preference
4. Validate with Survey

Best Practices

1. Use matched panels - Same personas evaluating both sides removes persona variance
2. Run sufficient sample - At least 100 for statistical confidence
3. Control variables - Compare one thing at a time
4. Segment results - Overall preference may hide segment-specific patterns

Related Articles

• Discovery - Initial audience research
• Surveys - Detailed quantitative follow-up
• Statistical Methods - Understanding significance