Key takeaways
- An AI-moderated interview is a research interview led by an AI moderator that asks questions, follows up on vague or interesting answers and records everything for analysis.
- There are five AI-moderated interview formats on Terapage: text, WhatsApp, voice, video and telephone (AI-powered CATI).
- Text, WhatsApp, voice and video interviews share one four-step set-up; telephone interviews add campaign, participant import, test-call and compliance steps.
- The right format depends on who your participants are and what you need to learn, from considered written answers to emotion, reactions and reach without internet.
- Every format feeds the same AI analysis: searchable transcripts, speech maps, sentiment and intensity, AI excerpts and quote analysis.
- AI has moved from experiment to everyday business practice. According to Amazon’s 2026 Unlocking Europe’s AI Potential research, 54% of European businesses now use AI, up from just 33% two years earlier. Market research is changing just as fast. Teams are expected to deliver deeper insight from more conversations, in more markets, in less time.
AI-moderated interviews meet that demand. One AI moderator can run hundreds of interviews at once, and each conversation keeps the depth of a one-to-one interview. But the format matters as much as the moderator: a busy shopper may prefer to reply on WhatsApp, while a rural respondent may only answer the phone.
This guide covers the five formats of AI-moderated interviews on Terapage: text, WhatsApp, voice, video and telephone. For each one, it explains when the format works best, how a study is set up and how the data is analysed.
What Is an AI-Moderated Interview?
An AI-moderated interview is a research interview led by an AI moderator. It is built on conversational AI, which IBM describes as technologies “such as chatbots or virtual agents, that users can talk to.” The AI moderator asks your questions, listens to each answer, follows up when an answer is vague or interesting, and records everything for analysis.
Instead of a fixed script, the AI moderator follows a discussion guide and responds to what each participant says, much like an experienced interviewer. New to the method? Start with this guide to conducting your first AI-moderated interview.
**
How Does an AI-Moderated Interview Work?**
Researchers set up the study once, and the AI moderator then runs every interview, asks follow-up questions and records each answer for analysis. Text, WhatsApp, voice and video interviews share the same four-step set-up, while telephone interviews have their own.
1. Choose the AI-Moderated Interview Format
Researchers start by choosing how participants will interact with the AI moderator: an interactive video call, an interactive voice call or a guided text and chat interview. Chat interviews can also run on WhatsApp.
Figure 1: Choosing the interview mode for an AI-moderated interview on Terapage.
Figure 2: Selecting WhatsApp Chat as the chat channel for a guided text and chat interview.
2. Add the Study Objective, Questions and AI Probes
Next, researchers describe the study objective and research theme, and AI can generate the interview questions from it. Each question can then carry follow-up probes, which researchers write themselves or let AI suggest, so the moderator asks for more detail whenever an answer is short or unclear.
Figure 3: Adding the activity details and study objective so AI can generate interview questions.
Figure 4: AI-generated interview questions, each with its own follow-up probes.
Figure 5: Adding an interview question with follow-up probes for the AI moderator.
3. Set Up the AI Moderator
Researchers choose the moderator’s name, tone of voice, speech voice and avatar once. Every participant then hears questions asked in the same way, which keeps interviews consistent and free from interviewer bias.
Figure 6: Setting the AI moderator’s name, tone of voice, speech voice and avatar.
4. Review and Launch the Interview
Finally, researchers check the activity overview, questions, AI moderator, schedule and participant incentives on one screen. Participants can come from the existing study, an insight community or participant recruitment, and the interview is published in one click.
Figure 7: Reviewing an AI-moderated interview before publishing, with the schedule and participant rewards.
What Are the 5 Formats of AI-Moderated Interviews?
Every format uses the same AI moderator, adaptive probing and analysis. What changes is how participants take part. On Terapage, they can join on any device, through the channel that suits them best.
1. Text-Based AI-Moderated Interviews
In a text-based interview, participants type their answers at their own pace, usually on their phone. The AI moderator reads each reply and follows up on it, for example by asking participants to walk through a real example instead of giving a general opinion. When the interview ends, responses are saved instantly for analysis.
Best for:
- Sensitive topics where participants prefer privacy
- Participants who like to think before they answer
- Studies that need clean written responses with no transcription
Figure 8: A text-based AI-moderated interview on mobile, from a follow-up probe to the completed screen.
Participants who prefer not to type can record their answer instead, and it is transcribed straight into the chat. Read more about what works in AI-moderated chat interviews.
Figure 9: A participant records an answer in a text-based AI-moderated interview, which is transcribed into the chat.
2. WhatsApp-Based AI-Moderated Interviews
In a WhatsApp-based AI-moderated interview, participants take part on a messaging app they already use every day. Researchers simply select WhatsApp as the chat channel during set-up. Participants can reply with text or voice notes, and the AI moderator follows up on each answer.
Best for:
- Consumer and mobile-first audiences
- Shopper and usage studies for consumer goods and services
- Markets where WhatsApp is the main way people communicate
Figure 10: An AI-moderated WhatsApp interview on mobile, with a participant replying by voice note.
Participants can also share photos and screenshots during the conversation, such as an error message from a failed checkout, and the AI moderator probes for more detail.
Figure 11: A participant shares a screenshot in an AI-moderated WhatsApp interview and the AI moderator probes further.
3. Voice-Based AI-Moderated Interviews
In a voice-based AI-moderated interview, participants speak naturally with the AI moderator, just as they would on a call. Every answer is transcribed live, so tone, hesitation and detail are captured without anyone taking notes. Learn how AI-led automated interviews are transforming conversational research.
Best for:
- Participants who find speaking easier than typing
- Emotional or experience-led topics
- Longer stories that lose detail in text
Figure 12: A voice-based AI-moderated interview on mobile, with a live transcript.
4. Video-Based AI-Moderated Interviews
In a video-based AI-moderated interview, the AI moderator appears as an avatar with real-time facial expressions. It listens to each answer and keeps a live transcript, much like a face-to-face interview. Researchers can preview the interview before launch. Our step-by-step guide to AI-moderated video interviews walks through the full set-up.
Best for:
- User experience research
- Concept, ad and product testing, such as online qualitative concept testing
- Studies where seeing reactions adds meaning
Figure 13: Previewing a video-based AI-moderated interview, with the AI moderator listening and a live transcript.
Participants can join from a laptop or a mobile phone, so they can take part from wherever they are.
Figure 14: A participant taking part in a video-based AI-moderated interview on mobile.
5. AI-Moderated Telephone Interviews (AI-Powered CATI)
In AI-moderated telephone interviews, the AI moderator calls participants at scheduled times and conducts the interview naturally, asking questions and following up just as it does in the other formats. This brings AI to traditional phone-based research without the need for a call centre.
*Best for:
*
- Participants who prefer phone calls
- Audiences with limited internet access
- Large-scale fieldwork with wide geographic reach
- Public health and healthcare research across regions
How to Set Up an AI-Moderated Telephone Interview
Create the campaign: Researchers choose the campaign purpose, such as market research or recruitment, and add the research activity name and description.
Figure 15: Setting the campaign purpose and research activity details for an AI-moderated telephone interview.
Add questions and probes: AI drafts the questions and follow-up probes from the study description, and researchers review, edit or add to them before any calls are made.
Figure 16: AI-generated questions and follow-up probes for an AI-moderated telephone interview.
Add participants: Participants can come from the current study, a CSV file or connected tools such as Salesforce, with phone numbers checked automatically for missing or duplicate entries.
Figure 17: Adding participants to an AI-moderated telephone interview campaign.
Figure 18: Importing participants from Salesforce for an AI-moderated telephone interview.
Preview the call: Before launch, researchers choose the AI interviewer’s voice and tone, then place a short test call to their own phone to hear the interview exactly as participants will.
Figure 19: Previewing an AI-moderated telephone interview with a test call to the researcher’s phone.
5. Compliance Built Into Every Call
Compliance is managed inside the campaign. Researchers can keep a Do-Not-Call list, track opt-outs and check country rules before calls go out, so every participant is contacted responsibly.
Figure 20: Managing the Do-Not-Call list, opt-outs and country rules for AI-moderated telephone interviews.
6. Tracking and Analysing AI Telephone Interview Results
Once calls begin, one dashboard tracks participants, total calls, completions, answer rate, average duration and opt-outs. Researchers can read every transcript, review key takeaways, transcript insights and a word cloud, and use Talk to your data to ask questions in plain language. Read more about how AI telephone interviews work, or how recruitment teams use them to screen candidates faster.
Figure 21: A results and AI analysis dashboard for AI-moderated telephone interviews.
How to Choose the Right AI-Moderated Interview Format
The right format depends on who your participants are and what you need to learn. Use this quick guide to match each format to your study.
Figure 22: When to use each of the five AI-moderated interview formats, and what participants need.
Some conversations still need a person, such as very sensitive interviews or stakeholder sessions. For those, many teams combine AI-moderated interviews with live interviews through hybrid research.
**
How Is Data From AI-Moderated Interviews Analysed?**
Analysis does not wait for fieldwork to end. AI-powered insights start working during the conversation and cover every format, so researchers can unlock insights with AI without coding answers by hand.
Transcripts and Speech Maps
Every interview has a searchable transcript, with an AI summary, AI analysis, word cloud and uploaded media in one place. Each participant answer is tagged, so researchers can quickly spot the most useful moments.
Figure 23: A searchable interview transcript with tagged participant answers and AI analysis tabs.
For group and spoken interviews, a speech map shows when each participant spoke and for how long, which helps researchers see who led the conversation.
Figure 24: A speech map timeline showing when each participant spoke and for how long.
AI Sentiment Analysis and Intensity
Responses are grouped into mood categories, such as positive learning or cautious optimism, and the granular sentiments behind them. Each mood shows its responses, share, average weight and strength. Learn more in our beginner’s guide to AI sentiment analysis.
Figure 25: Sentiment analysis of AI-moderated interviews, with mood categories and intensity.
The same results appear in a sentiment table that researchers can filter by emotion and strength, sort and export.
Figure 26: A sentiment table linking each emotion to a key participant statement.
AI Excerpts and Quote Analysis
The AI groups what participants said into topics and ranks them by the number of responses. Hovering over any topic reveals the verbatim quotes behind it.
Figure 27: AI excerpts ranking interview topics, with the verbatim quotes behind each one.
Quote analysis then ranks themes by how often they were mentioned, with every source quote one click away. Findings can move straight into reports and dashboards or branded insight reports.
Figure 28: Quote analysis ranking themes by how often participants mentioned them.
Consent, Security and AI Transparency in AI-Moderated Interviews
Consent is part of every study set-up. Before an interview begins, participants review and accept the study terms, such as a confidentiality agreement, through a mandatory e-consent. All responses are stored on a secure, GDPR-compliant platform, in line with Terapage’s AI transparency principles.
Figure 29: A mandatory e-consent that participants accept before an AI-moderated interview.
**
Choose the AI-Moderated Interview Format That Fits Your Participants**
The best AI-moderated interview is the one participants are comfortable taking part in. Text suits considered answers, WhatsApp suits everyday engagement, voice and video capture emotion and reaction, and telephone reaches people that online methods miss. On Terapage, all five formats share one set-up, one AI moderator and one analysis workspace, alongside every other platform feature for in-house teams and research agencies.
Figure 30: Terapage’s five research products and shared platform features on one unified platform.
To try AI-moderated interviews on your next study, start a free trial or request a demo.
Top comments (0)