The level of AI Moderator follow-up for an AI Moderated Question is guided by your selected follow-up depth as well as your study goals. While study goals is optional, we recommend using this as a best practice to help guide the AI Interviewer on follow-up questions.
The Goal field appears when Probe depth is set to Probe lightly for clarity or Follow up for richer detail. It gives the AI moderator guidance on what follow-up questions should aim to uncover.
Participants never see or hear the Goal. They only hear your question text.
What the Goal does
The Goal is added into the instructions for the AI moderator as background context. It helps the AI decide what direction to take when it asks a follow-up — not how many follow-ups to ask, and not what question to read aloud.
Setting | What it controls |
Question text | The exact question read aloud to the participant |
Probe depth | How many follow-ups the AI may ask |
Goal | What themes or missing details follow-ups should explore |
The AI will obey probe depth limits and built-in stop conditions. A detailed Goal cannot override those rules.
What belongs in a Goal
What you need to learn from this answer
Topics to probe if the participant does not mention them
What a "good enough" answer looks like
1–3 short, actionable sentences
Best practices
1. Keep it short (1–3 sentences)
Aim for under 300 characters when possible. The hard limit is 2,000 characters. The Goal provides guidance; it is not a script.
Example: Understand what an ideal checkout flow involves. If they mention multiple pain points, ask which one is the biggest pain point. If they mention specific issues, probe for more detail, such as what step it occurs in.
Avoid: A multi-paragraph briefing with timing rules, conditional branches, topic lists, and moderation instructions.
2. Describe outcomes, not a conversation script
Write what you want to learn, not what the AI should say or do step by step.
Avoid:
Timing instructions ("Spend no more than X seconds")
Follow-up counts ("Ask a maximum of 1 follow-up question")
Long conditional trees ("If they mention X, ask Y. If they mention Z, ask W…")
Meta framing (Coming in full Ai mod Interviews release) ("You are moderating a pilot kickoff survey…", "The survey is being run by…")
Tone or style rules that duplicate built-in behaviour (Coming in full AI Interviewer release) ("Avoid turning this into a product pitch")
These can conflict with the AI Moderator settings and make follow-ups unfocused or inconsistent.
3. Do not repeat or replace the question text
The AI reads the survey question exactly as written. The Goal should add research intent that the question alone does not convey.
If your Goal reads like a second question or a full interview plan, move that content into the question text or simplify the Goal to a single learning objective.
4. Focus on 1–2 priority themes, not a keyword list
"Listen for…" lists dilute follow-ups. The AI may chase every topic instead of going deep on what matters.
Instead of listing many product capabilities, you can focus on prioritized themes.
Example: Learn their top priorities and decision criteria for adopting a research tool. Probe for the study types and outputs that matter most to their team.
5. Define "sufficient" so the AI knows when to stop
This helps the AI accept good answers instead of over-probing.
Example: A sufficient answer includes their top 2–3 priorities and why those matter. If they give a clear opinion with a reason or example, do not keep probing.
6. Use "if missing, probe for…" sparingly
One or two "do not mention" areas works well. More than that usually belongs in separate questions.
Best practice example: If they do not mention analysis needs, ask what outputs would be most useful for their team.
Avoid too many references.
Example: Six separate "if they mention X, ask Y" rules in one Goal.
7. Match Goal ambition to probe depth
Probe depth | Goal should… |
Collect answer & move on | Goal field is hidden — not used |
Probe lightly for clarity | Stay narrow: clarify vagueness or missing reasoning |
Follow up for richer detail | Can mention 2–3 themes to explore, but still keep it brief |
A detailed Goal with Probe lightly will still only get at most 2 follow-ups. Do not write Goal text that assumes a full qualitative interview.
Example Goals by question type
Exploratory
Understand their current workflow and where they lose the most time. If they only describe tools, ask what feels most tedious or manual.
Evaluation / priorities
Learn their top 2–3 requirements for purchasing and why those matter. If they mention privacy or security, probe for what would build stakeholder confidence.
Warm-up / rapport
Build context on their role and how this workflow fits into their team. A sufficient answer includes their role and what a typical week looks like.
Quick checklist
Before saving the Goal, ask:
Is this 1–3 sentences?
Does it say what I need to learn, not how to run the interview?
Did I leave timing and follow-up count to probe depth?
Are there 1–2 probe hints, not a branching decision tree?
Would I be happy if the AI used this to ask one focused follow-up?
For the question text: Is this text what you want participants to hear?