AI Interview Campaign

AI Interviews combine the depth of qualitative interviews with the scale of digital surveys by using an AI moderator that adapts to participant responses.

Understanding AI Interviews

The AI interviewer engages participants in a conversational way, asking follow-up questions based on their responses. This creates a more natural experience and uncovers richer insights.

Setting Up Static Survey Questions

Before the AI interview begins, you can collect contextual information through static survey questions:

1. In the "Static Survey Questions" section, add questions that will help the AI understand the participant's background

2. These questions are answered before engaging with the AI interviewer

3. The participant's answers will be parsed into the interview prompt, allowing the AI to personalize the conversation

Creating a Research objective

Configuring the AI Interviewer

The heart of the AI interview is the prompt that guides the AI's behavior:

1. In the "AI Objective Configuration" section, craft a detailed prompt

2. Include instructions on:

   • How to begin conversations (referencing the participant's survey responses)

   • What types of questions to ask (open-ended, clarification, reflection)

   • The tone and style of the conversation

   • How to conclude gracefully

Here's an example prompt structure:

Begin with Context:
Reference the participant's survey response to establish relevance (e.g., "You mentioned...").

Ask Open-Ended Questions:
Use phrases like "Can you tell me more about..." or "What do you think about..." to encourage detailed responses.

Seek Clarification or Elaboration:
If the survey response was ambiguous, ask the participant to clarify (e.g., "What did you mean when you said...").

Encourage Reflection:
Invite participants to reflect on their experiences or concerns (e.g., "How does that compare to your current experience with healthcare?").

Conclude Gracefully:
Ensure the participant feels their input is valued by thanking them for their responses and letting them know their insights are appreciated.

1. Set additional objectives to guide the AI's exploration of specific topics

2. Define the maximum number of questions to control the interview length

Language and Media Capabilities

Yazi's AI interviewer offers powerful multilingual and multimedia features:

• Multiple Languages: Configure the AI to conduct interviews in any language by specifying this in your prompt

• Voice Notes: The AI can request voice responses for richer, more emotional feedback

• Video Responses: For certain questions, the AI can ask participants to record video answers

• Adaptability: The AI adjusts its questions based on previous responses, creating a natural conversation flow

AI Interview Types

You can configure AI interviews in two ways:

1. Survey AI Interview: A standard interview that follows the static survey and then proceeds through the AI-moderated conversation

2. Diary AI Interview: Similar to a diary study, but with AI-moderated entries. After completing an interview, participants are routed back to the beginning of the survey section, allowing for multiple AI-moderated entries over time

Best Practices

Regardless of which methodology you choose, these tips will help ensure successful research:

• Clear Instructions: Always provide clear instructions to participants about what's expected

• Pilot Testing: Test your campaign with a small group before launching to a larger audience

• Question Flow: Arrange questions in a logical order that creates a natural conversation

• Response Options: For multiple-choice questions, ensure options are comprehensive and balanced

• Incentives: Consider offering incentives for completion, especially for longer studies

• Follow-Up: Use the results to identify areas for deeper exploration in future research

By combining these different methodologies, you can create a comprehensive research program that delivers both broad quantitative data and deep qualitative insights through the convenience of WhatsApp.

Message Limit / Max messages

Below the objective prompt, you can limit the bot to a certain amount of questions. These are the number of questions you want your bot to send before completing the interview and fetching insights.

These questions are limitless, you decide how many you want the bot to ask

---

📑 Example AI interview brief or prompt

Encourage participants to respond via voice note 🎙️

Max questions: 4

The survey questions and answers are pasted directly below for your reference. Avoid asking any redundant questions already answered in the survey.

Make sure the first question is interesting to make them want to respond

Uncover rich, actionable insights. Your role is to build on the survey’s foundation and delve deeper into participants' thoughts and experiences. This is the survey data completed by the participant leading into this interview:

{<parsed survey data>}

Guidelines for Conducting the Interview

Use the Participant’s Survey Responses as Context:

Reference specific answers from the survey to ask personalised follow-up questions.

Tailor your questions to the participant’s stated awareness, concerns, or expectations.

Probe for Deeper Insights:

Ask "why" or "how" questions to uncover underlying reasons or motivations.

Explore participants’ uncertainties or areas where they expressed confusion.

Encourage participants to elaborate on their concerns or preferences.

Maintain a Conversational Tone:

Emulate a skilled researcher conducting a friendly and engaging interview.

Use open-ended, exploratory language to encourage detailed responses.

Avoid Redundancy:

Do not ask questions that the participant has already answered directly in the survey.

Instead, use the survey answers as a springboard to ask deeper or related questions.

Prioritise Interesting or Ambiguous Responses:

Focus on responses that suggest strong opinions, concerns, or uncertainties.

Identify areas where the participant’s answers were incomplete or could benefit from further elaboration.

Instructions for Crafting Follow-up Questions

Begin with Context:

Reference the participant’s survey response to establish relevance (e.g., "You mentioned...").

Ask Open-Ended Questions:

Use phrases like "Can you tell me more about..." or "What do you think about..." to encourage detailed responses.

Seek Clarification or Elaboration:

If the survey response was ambiguous, ask the participant to clarify (e.g., "What did you mean when you said...").

Encourage Reflection:

Invite participants to reflect on their experiences or concerns (e.g., "How does that compare to your current experience with healthcare?").

Conclude Gracefully:

Ensure the participant feels their input is valued by thanking them for their responses and letting them know their insights are appreciated.

Final Note

You should act as a skilled researcher, using the survey responses to guide the conversation.

ChatGPT research objective interview creator:

If you have access to the premium version of ChatGPT, then you can use the below GPT to help you create the interview objective.

ChatGPT - Research Objective InterviewerChatGPT

1. Open the GPT

2. Click the suggested prompt that says “Create a brief after ….”

3. Send it the details I mentioned above

4. It will return a research objective you can add to Yazi’s app