What Are AI Follow-Ups?
When enabled, the AI analyzes participant responses in real-time and generates relevant follow-up questions. This mimics how a skilled interviewer would probe for more detail. Example:Original question: “What do you like most about this product?” Participant response: “I love how easy it is to use, especially the quick setup process.” AI follow-up: “You mentioned the quick setup process. Can you describe what made it feel quick and easy for you?”
Benefits of AI Follow-Ups
Deeper Insights
- Probe into interesting responses automatically
- Uncover the “why” behind answers
- Get specific examples and details
Consistent Quality
- Every participant gets appropriate follow-ups
- No interviewer fatigue or inconsistency
- Systematic probing across all responses
Scalability
- Get interview-quality data at survey scale
- No need for live interviewers
- Consistent probing for hundreds of participants
How AI Follow-Ups Work
1
Participant Answers
The participant responds to a question (typically open-ended or video/audio).
2
AI Analysis
Deepfield’s AI analyzes the response for:
- Key themes and topics mentioned
- Interesting claims or opinions
- Opportunities for clarification
- Gaps that need exploration
3
Follow-Up Generation
The AI generates a relevant, contextual follow-up question based on what the participant said.
4
Participant Responds
The participant sees the follow-up and provides additional detail.
5
Optional: More Follow-Ups
Depending on settings, additional follow-ups may be generated.
Enabling AI Follow-Ups
1
Select a Question
Click on the question where you want AI follow-ups.
2
Open Advanced Settings
Find the AI follow-up settings in the question panel.
3
Enable Follow-Ups
Turn on the AI follow-up feature for this question.
4
Configure Options
Set the number of follow-ups and any custom instructions.
5
Save Changes
Save your study to apply the settings.
Configuration Options
Number of Follow-Ups
Control how many follow-up questions the AI generates:| Setting | Use Case |
|---|---|
| 1 follow-up | Light probing, shorter surveys |
| 2 follow-ups | Standard depth |
| 3+ follow-ups | Deep exploration, interview-style |
More follow-ups = deeper insights but longer participant time.
Custom Instructions
Provide guidance to the AI on how to probe: Example instructions:- “Focus on emotional reactions and feelings”
- “Ask for specific examples whenever possible”
- “Probe into price-related concerns”
- “Explore comparison with competitors”
Follow-Up Triggers
Choose when follow-ups should occur:- Always: Every response gets follow-ups
- Conditional: Only when certain themes are detected
- Length-based: Only for shorter responses that need elaboration
Best Practices
Choose the Right Questions
AI follow-ups work best with:- Open-ended text questions
- Video and audio responses
- Questions about experiences, opinions, or preferences
- Multiple choice questions
- Simple factual questions
- Yes/no questions
Write Good Initial Questions
The quality of the initial question affects follow-up quality: Good: “Tell me about your experience using our mobile app for the first time.” Less effective: “Do you like our app?” Open, exploratory questions give the AI more to work with.Set Appropriate Depth
Match follow-up depth to your research needs:| Research Type | Recommended Follow-Ups |
|---|---|
| Quick pulse survey | 1 or none |
| Standard research | 1-2 |
| Deep qualitative | 2-3 |
Provide Context via Instructions
Help the AI understand your priorities:“This study is about understanding barriers to purchase. When probing, focus on concerns, doubts, or hesitations the participant mentions.”
What AI Follow-Ups Ask About
The AI typically probes for:| Topic | Example Follow-Up |
|---|---|
| Clarification | ”What do you mean by ‘confusing’?” |
| Examples | ”Can you give me a specific example?” |
| Reasons | ”Why do you feel that way?” |
| Comparisons | ”How does this compare to alternatives?” |
| Emotions | ”How did that make you feel?” |
| Specifics | ”When exactly did that happen?” |
AI Follow-Ups in Practice
Example Flow
Q1: “What was your first impression when you saw our new packaging?” Response: “It looks more premium now. The colors are better.” AI Follow-Up 1: “You mentioned it looks more premium. What specific aspects of the design give you that impression?” Response: “The matte finish and the gold accents make it feel more expensive.” AI Follow-Up 2: “How important is this premium look when deciding whether to purchase this type of product?” Response: “Pretty important actually. I associate better packaging with better quality inside.”What You Learn
From this exchange, you learn:- The redesign successfully conveys premium positioning
- Specific elements (matte finish, gold accents) drive perception
- Packaging influences quality expectations and purchase decisions
Viewing AI Follow-Up Data
In your responses, you’ll see:- The original question and response
- Each follow-up question and response
- All exchanges as a conversation thread
- AI analyzes all responses including follow-ups
- Insights incorporate the deeper context
- Citations may reference follow-up exchanges
Common Mistakes
Too many follow-ups everywhere
Too many follow-ups everywhere
Don’t enable follow-ups on every question. This makes surveys too long. Choose strategically.
Using with wrong question types
Using with wrong question types
Follow-ups work poorly with multiple choice or simple questions. Use with open-ended questions.
No custom instructions
No custom instructions
The AI does a good job by default, but custom instructions aligned with your research objectives produce better probing.
Not testing
Not testing
Test the follow-up experience yourself. Make sure follow-ups are relevant and add value.
AI Follow-Ups and Survey Length
Each follow-up adds time to the participant experience:| Base Question | + 1 Follow-Up | + 2 Follow-Ups |
|---|---|---|
| 1 minute | 2 minutes | 3 minutes |