Best practice guide for supporting AI-Moderated Interviews

We encourage supply partners that want to effectively monetize AI-Moderated Interview opportunities to review this document and their integration.

Unlike traditional quantitative surveys, AI-Moderated Interviews (AIMI) represent a ‘Qual-at-Scale’ methodology. AIMI uses conversational AI to conduct one-on-one, open-ended, voice and/or video-enabled interviews at scale. AIMI allows for real-time probing  (i.e., simulating a human moderator) across hundreds of participants simultaneously.

This methodology necessitates a higher level of technical readiness and deeper respondent engagement than standard studies.

The demand for users to participate in AI-Moderated Interviews has seen significant growth since 2025 and is projected to continue expanding as the technology advances.

This increase is driven not only by new technology but also by research, such as that conducted by the University of Mannheim, which indicates that AI-Moderated Interviews offer superior linguistic quality to traditional quantitative surveys. Specifically, respondents use 51% more unique words and discuss 36% more unique themes, establishing this as a scalable and in-depth research method.

AI-Moderated Interviews are typically characterised by:

• Higher CPIs offered to reflect the increased effort by the respondent

• Increased study LOI, which also leads to higher CPIs offered

• Increased conversion rates as the participants tend to be more highly engaged and better rewarded for their time and effort

AI-Moderated Interviews present an excellent opportunity to increase conversion and revenue rates and boost respondent engagement and retention. The demand for these types of surveys is expected to increase as buyers experiment and scale this methodology.

AI-Moderated Interviews are typically run by:

• Customers who need human qualitative data at scale, for example, to validate product concepts or to understand brand preferences.

• Brands, political pollsters, media, or research agencies running AIMI studies via video, audio, or text.

• Project managers that need to run studies that require higher respondent effort and variability than traditional quantitative surveys.

• Project managers who need to run studies where traditional survey benchmarks, e.g., Length of Interview (LOI), Incidence Rate (IR), and Cost per Interview (CPI), may not directly apply.

Similar to traditional survey research methods, we see a range of target audiences requested by AIMI buyers.

1. Enable Opportunities that collect PII  

Supply partners leveraging our programmatic integrations must first ensure they are not excluding opportunities that collect PII to be able to sample to AIMI survey opportunities.

• Feed Partners:

  • Confirm your subscription is not excluding PII opportunities. Review your current Feed Opportunities subscription settings to ensure your subscription does not include a filter on PII collection. If your subscription includes in the body the property collects_pii set to FALSE, your system is not receiving PII surveys.  You will need to reset your subscription and either change this to TRUE or omit the argument entirely.

  • Confirm your survey evaluation logic does not exclude PII opportunities. If your subscription does not exclude PII opportunities, also confirm that any rules for scoring or selecting surveys for respondents does not exclude or demote surveys where collects_pii is TRUE.

  • PII studies can be identified via the collects_pii property in the Feed Opportunities response.

• Match Partners:

  • Including the allow_pii_surveys=true parameter on entry links enables PII opportunities. More information can be found under “Entry Links” in our Match Integration Guide.

2. Map to AIMI Qualifications

To ensure you can present audio and video surveys in your offerwall, please ensure you have mapped your integration to the following qualifications:

Please review this document to access the full Cint Standard Qualification Library and read the latest updates.

3. Respondent Criteria Recommendations

As AIMI studies require increased focus and effort from respondents, we recommend prioritizing proven, high-quality respondents. This may include:

• Excluding respondents who have not previously interacted with any surveys

• Prioritising respondents who have previously participated in video, audio, or extended LOI surveys successfully

4. Respondent Experience Recommendations

Video and Audio labels (Feed API only)

We recommend displaying labels for audio or video requirements when offering an AIMI opportunity to a respondent.

This may include an icon of a webcam or a microphone or text labels such as ‘Video’ or ‘Audio’.

As respondents will be more easily able to identify that it’s a conversational study, we expect that there will be a lower drop-off rate, as participants who are unwilling to participate or are not in a quiet environment will not enter the study.

Examples:

Video survey

Audio survey

Pre-survey notification (Feed API and Match)

We recommend that partners add a pre-survey notification to alert respondents that the study will require a quiet environment to provide responses in audio or video format.

By adding this notification to the respondent journey prior to survey entry, respondents will be clearly informed about the requirements to participate and have the opportunity to exit the survey without registering as a drop-off and whilst maintaining their engagement with other available surveys.

For Match Supply Partners: when enabling PII for the respondent session, you should prompt the user that they may enter a survey that requires audio or video responses.

Example:

The information supplied by Cint will show whether the study requires audio or video responses, so you can reliably display the correct information in the offerwall.

AIMI studies are prone to higher reconciliation rates if effective controls are not implemented by both the buyer and supplier. Suppliers can help reduce reconciliation rates by:

• Inform: Let panelists know that they are entering a study involving audio or video recording by utilizing the previously mentioned labels and alerts. Ideally, implement verification checks to confirm the respondent's audio and video functionality prior to survey entry.

• Educate: Ensure panelists understand standards for audio and video studies such as:

  • Remaining fully visible on camera at all times

  • Answering honestly and giving detailed feedback

  • Dressing appropriately the duration of the interview

  • Refraining from the usage of foul or offensive language

  • Finding a quiet environment where they are unlikely to be disturbed

• Verify: Put in place strong verifications and protective measures throughout the panelists entire journey from recruitment and onboarding to payouts. The  Global Data Quality (GDQ) Initiative has produced an Internal Approaches document for a full list of potential measures that can be taken to improve data quality of panelists.

Suppliers can utilize the comprehensive Safeguarding Qualitative Research guide, developed by the GDQ to assist qualitative researchers, to better understand researcher expectations and panelist considerations. Our primary recommendations, derived from this resource, include the following:

1. Verify: Qualitative researchers may challenge the identity or persona of the individuals they have interviewed if they have quality concerns. Being able to assure researchers that the panelists identity and profiling is verified and accurate can help reduce reconciliations.

   • Questions To Consider: Has the panelist shown themselves to be a real person? Can the identity they have registered with be proven as true?

   • Top Tip: Verifying identities must be handled with caution due to privacy laws and sensitivities within certain geos or demographics. Consider utilizing specialized third party verification services to help with this.

2. Identify: Standards within qualitative research are often higher as issues with respondents are more easily identifiable than in quantitative research where a fraudulent or poor quality response could remain undetected in aggregated numeric datasets. Being able to assure researches that panelists are high quality survey takers can help reduce reconciliations:

   • Questions To Consider: Is the panelist tenured with an understanding of how surveys work? Or are they a new panelist who may not understand the expectations of AIMI projects? Can you provide evidence of their survey history and quality measures they have passed?

   • Top Tip: The Cint Exchange requires buyers to provide reasons for reversals. Cint provides suppliers with access to these reversal reasons which can be used to enrich their understanding of panelists quality.

Please reach out to your Account Team if you have any questions about monetizing audio and video surveys.