Understanding AI Moderated Interviews
Unlike traditional, static interviews, AI Agent–led interviews are not fully deterministic.
The AI adapts dynamically to participant responses - following the project plan in most cases but occasionally shifting focus based on context.
This adaptability makes Genway powerful. However, it also means your project details must provide clear structure and context to help the AI deliver consistent, meaningful results.
Objectives: Setting Clear Intentions
Your project objectives anchor the AI Agent’s overall direction.
If objectives lack context, the AI may ask generic or repetitive questions, shorten the conversation prematurely, or fail to extract useful insights.
Example Objective:
“Understand how small business owners choose delivery partners for local orders.”
This gives the AI a clear mission - focusing on decision-making factors, pain points, and desired outcomes.
If the objective were vague - e.g., “Learn about delivery services” — the AI would lack focus, leading to surface-level insights.
Pro Tip:
Always explain the “why” behind your research. Context helps the AI connect your project’s purpose with the participant’s real-world experience.
Niche Use Case Example
When your project explores a new or highly specific topic, the AI Agent might not have enough prior context to guide the conversation effectively.
Example: If you’re studying behaviors related to a new subscription plan of an AI tool, the AI Agent may not yet understand the plan’s unique features, pricing tiers, or use cases. As a result, it might struggle to navigate the discussion naturally or follow up on key points.
To ensure accuracy and flow:
Include detailed descriptions of the plan in the project description and objectives.
Add targeted example questions in the Topics section that reference specific characteristics of the new plan.
Example question:
“How does the adaptive learning feature in your Pro subscription affect your workflow?”
Reinforce these details in the Guidelines section to help the AI maintain focus and relevance.
Providing rich context ensures that the AI can conduct confident, meaningful interviews - even in specialized or unfamiliar domains.
Topics: Structuring the Conversation
Topics define what the AI should explore during the conversation.
Each topic can include example questions - but too many can overload the AI, causing it to skip important sections due to time limits.
Best Practices
Keep 2–3 well-chosen example questions per topic, and avoid exceeding a total of 10–12 questions overall across all topics to maintain a balanced and focused interview.
Arrange questions logically if they depend on each other.
Avoid stacking conditional questions — these are follow-up questions that depend on specific answers (e.g., “If yes, then ask…”). The AI already manages such branching dynamically, so you don’t need to include them explicitly..
Include one or two probing instructions, like:
“Ask follow-up questions to understand the ‘why’ behind participant choices.”
Example Topic: Evaluating Digital Advertising Tools
Example Question | Purpose |
How do participants choose where to advertise their products? | Understand decision criteria |
What role does budget play in their decisions? | Assess trade-offs |
What challenges do they face when tracking campaign success? | Identify pain points |
⚠️ Note: If too many sub-questions are added, the AI may rush or omit sections to stay within time limits.
Project Outcomes: Guiding the Depth of Insight
Project Outcomes help the AI control the depth and tone of questioning.
If outcomes are framed too broadly, the AI may not know when to probe deeper or when to move on.
Example Outcome:
“Identify actionable insights about pain points in the onboarding process.”
You can reinforce this with a guideline such as:
“Prioritize actionable insights by asking specific questions about SMB pain points and feature opportunities, without compromising on any topic from the project details.”
⚠️ Note: Guidelines like this deepen the AI’s analysis but may also extend interview length and potentially make the AI Agent rush over other topics. Always test before publishing to confirm pacing and topic balance.
Audience: Defining Who the AI Talks To
The Audience section determines how the AI frames its language, tone, and examples.
Example Audience:
“Small business owners using online sales tools to reach new customers.”
This definition helps the AI adapt its tone - professional but approachable, and choose relevant examples.
If defined vaguely (e.g., “business users”), the AI might use irrelevant terminology or skip key context.
Guidelines: Steering the AI’s Behavior
Guidelines have a significant impact on the interview flow - even small changes can strongly influence how the AI conducts and sequences the conversation. Keep your guidelines clean and concise, focusing only on the instructions that truly matter to the research goals.
⚠️ Note: It is not necessary to instruct the AI Interviewer to avoid asking Yes/No questions, leading questions, double-barreled questions, or interrupting participants. This is already managed automatically in the background as part of Genway’s intelligent moderation design.
Guidelines shape how the AI moderates.
While Genway’s AI Agents follow solid research principles by default, custom guidelines make interviews more precise.
Useful Examples:
“At the beginning of the interview, ask participants how long they’ve been using [product].”
“Keep discussion focused on feature usability rather than pricing.”
“Avoid technical details unless the participant brings them up.”
These help the AI maintain focus, consistency, and tone alignment with your research goals.
Testing and Iteration
Before publishing, always test your project setup.
Run a sandbox demo interview side-by-side with your project Topics to confirm that:
The AI follows the expected flow - topic by topic
Topic transitions are smooth
The interview ends naturally, not abruptly
If issues arise, adjust your objectives, topics, or guidelines, and test again.
Each iteration sharpens the AI’s understanding of your intent.
Designing Dual Interview Flows
If needed, you can design your interview to follow two paths based on whether one specific condition is met. You’ll need to ask the AI to begin the interview with a question that determines the participant’s path for the rest of the session, and make sure your project details clearly describe both paths from start to finish, as shown in the example below.
Example
In your objectives, clarify that the interview includes two flows and define what those flows are. Ensure that the topics section is clearly divided into Flow A and Flow B to reflect this structure.
2-Flow Objectives
The interview includes two flows based on user plan type:
Flow A - Standard Users
Explore core product experience
Identify challenges and upgrade motivators
Flow B - Premium Users
Explore premium feature value
Assess ROI and satisfaction drivers
2-Flow Topics
Before main questions, the AI asks:
“Are you currently using the Premium plan?”
If Yes then Follow Flow B (Premium Users)
If No then Follow Flow A (Standard Users)
Then, customize the topics:
Flow A | Flow B |
Feature usage | Premium feature value |
Support needs | ROI and retention |
Upgrade motivators | Advanced engagement |
This branching design allows one project to serve multiple user segments while preserving interview depth and quality.
2-Flow Guidelines
Begin the interview by asking if the participant is using the Premium plan so you can decide which flow to follow A or B
Final Thoughts
AI Moderated Interviews combine structure and adaptability.
By crafting thoughtful Objectives, Topics, Outcomes, and Guidelines, you empower your AI Agent to conduct rich, contextually aware conversations, delivering insights that go far beyond traditional research.
Thank you for following this guide, and happy crafting!