Automated Text Analysis
Reduce manual coding work on open-ended responses. Automated analysis suggests themes, sentiment, and candidate quotes that researchers can review and refine.
Available on Professional and Enterprise plans
Review-Assisted Analysis
Keep the repetitive coding work structured while preserving researcher judgment
What Automated Analysis Can Do
Theme Extraction
Identify recurring themes and topics from open-ended responses for researcher review
Example: Sample output: themes grouped from product feedback responses
Reduce manual coding work while keeping review in the loop
Sentiment Analysis
Classify emotional tone as positive, negative, neutral, or mixed
Example: Sample output: sentiment distribution with confidence indicators
Use sentiment as a starting point for review
Quote Discovery
Surface candidate verbatim quotes for each theme
Example: Sample output: quotes grouped under onboarding friction
Find candidate quotes without treating them as proof by default
Trend Detection
Track how themes and sentiment change over time across survey waves
Example: Sample output: performance mentions increased across recent waves
Flag emerging issues for follow-up analysis
See It in Action
NPS Comment Analysis
A representative respondent quote appears here for researcher review before it is used in reporting.
Popular Use Cases
Customer Feedback Analysis
Understand what customers love and hate about your product from NPS comments
Support Ticket Mining
Identify common pain points from support conversations and tickets
Employee Survey Insights
Extract actionable themes from engagement survey open-ends
Market Research
Analyze focus group transcripts and in-depth interview responses
Governed Analysis Controls
Workspace controls and provenance for sensitive survey analysis
Model-Use Terms
Confirm model-use terms for your workspace before sensitive analysis
Retention Controls
Retention and deletion workflows can be governed by workspace policy
Human Review
Suggested themes, sentiment, and quotes remain reviewable before reporting
8 AI data-handling controls are tracked in the trust registry.
Review AI data-handling postureFrequently Asked Questions
How accurate is the automated analysis?
Automated analysis is a review aid, not a replacement for researcher judgment. Use confidence indicators, inspect representative responses, and manually review high-stakes themes before making decisions.
What languages are supported?
Automated analysis currently works best for English. Broader language support (Spanish, French, German, Portuguese, Italian, Dutch, and Japanese) is in beta and expanding. For other languages, you can export data for external analysis.
Is my data used to train any models?
Automated analysis is designed for governed workspace workflows. Model-use terms, retention, access controls, audit logging, and deletion workflows should be confirmed for your workspace or enterprise agreement before processing sensitive data.
Can I customize the themes?
Yes. You can start with suggested themes and then refine them: merge similar themes, rename for your terminology, add custom themes, or remove irrelevant ones. The system will then re-classify responses based on your customizations.
Go Deeper
Research Guides Library
Practical guides on MaxDiff, conjoint, pricing research, NPS, and survey design.
Pricing Research Suite
Automated text analysis ships in Professional alongside conjoint, MaxDiff, and Van Westendorp.
Data Analysts
How analysts combine automated theme suggestions with their own review and exports.
Ready for Review-Assisted Analysis?
Structure open-ended responses faster while keeping researcher review in the workflow.