Overview
The Smart Probe Text Card is an AI-powered question type that automatically generates context aware follow up questions based on a respondent’s open-ended answer. It helps researchers uncover deeper insights by dynamically probing for more detail when responses are brief or incomplete. At the end of the probing sequence, Smart Probe also generates a concise summary that captures the key points from the conversation.
Why Use Smart Probe?
Smart Probe helps you:
Collect richer qualitative insights without adding more manual question writing.
Improve data depth and quality through context aware follow up questions.
Minimize respondent fatigue by only probing when necessary.
Enhance downstream AI analysis processes, like AI Insights Summarizer and Insight Reels, through stronger input quality.
Gain more actionable insights as Smart Probe helps uncover the “why” and “how” behind participant feedback.
How to Set Up a Smart Probe Text Card
1. Add the Smart Probe Text Card
In your chat flow, click the “+” icon.
Under Question Cards, select Smart Probe Text.
A new card (for example,
Smart_Probe_Text_1) will appear in your flow.Click Settings to open configuration options.
2. Configure the Topic Question
In the Probing tab, enter your main topic question. This is the core question you want respondents to answer.
This topic question acts as the anchor for the AI to generate relevant follow ups.
Click Add Another AI Generated Question to create multiple AI generated follow up questions within the same thread.
3. Set Display Logic with Thoughtfulness Score
Each follow up question can be shown conditionally based on the thoughtfulness score of the previous answer. The thoughtfulness score is an internal 1–10 rating assigned to the respondent’s answer based on 10 quality metrics such as relevance, clarity, completeness, and level of detail. Lower scores indicate shallow or vague responses, helping Smart Probe decide when a follow up is needed.
Toggle Enable Display Logic on.
Choose when to ask each follow-up:
Ask this question if the thoughtfulness score for the topic is lower than 7/10.
or Always ask this question.
4. Define Conversational Goals
In the Conversational Goals tab:
Describe what you want to understand from respondents’ answers.
Goals help ensure follow up questions align with your research intent.
Keep goals short and specific.
5. Add Avoidance Instructions
In the Things to Avoid tab:
Specify topics, tones, or question types that the AI should avoid.
This prevents redundancy or off-topic probes.
How to View Results
1. In Topline Reports
The Topline Report automatically includes an AI-generated summary of each Smart Probe thread.
This summary consolidates all follow up responses into a single narrative insight.
2. In Custom and Insights Reports
You can analyze Smart Probe data in Custom Reports and Insights Reports.
When you add a Smart Probe question to a dataset, all data points are included by default, allowing flexibility in what you choose to display.
3. In Chat Data
In Chat Data, only the Smart Probe summary is visible.
This allows quick access to high level insights without showing each probing exchange.
4. In Chat Exports
Chat Exports include the full probing history for each respondent, including:
The main topic question and initial response
All AI-generated follow up questions and responses
Thoughtfulness scores for each stage
The final summary
This provides a complete view of the respondent’s conversational journey for deeper analysis.
Best Practices
Limit each Smart Probe to one topic for clarity and consistency
Add specific conversational goals to help the AI stay aligned with your research intent.
Use avoidance instructions to keep probes focused.
Additional Notes
The Smart Probe Text Card can be used in any language, but it has been primarily trained and tested in English, and this is where performance is most reliable.
If you’re using Smart Probe in other languages, we recommend taking extra care to review the generated follow ups, thoughtfulness scoring, and summaries in a test chat before deploying to respondents.
Smart Probe is available only on domains with AI features enabled.
AI Models and Prompt Versions
Smart Probe uses the following GPT models and prompt configurations for question generation, response scoring, and summarization.
AI Function | GPT Model | Prompt Version | Release Date | Key Purpose / Benefits |
Generate Question | GPT-4.1 Mini | 1.0.0 | 18 September, 2025 | Generates context aware follow up questions based on respondent answers and author defined conversational goals. |
Get Thoughtfulness Score | GPT-4.1 Mini | 2.0.0 | 30 October, 2025 | Evaluates response depth and completeness using contextual understanding, ensuring accurate scoring aligned with the topic’s intent. |
Summarize Messages | GPT-4o | 1.0.0 | 18 September, 2025 | Produces a concise summary of the entire probing thread, combining multiple responses into a single, coherent insight. |