Consumer research has a trade-off. Surveys can reach thousands, but most answers lack depth. Human interviews provide the “why,” but they take time, cost money, and are hard to run regularly.
AI-moderated interviews could reduce this gap. AI can conduct interviews faster, and consumers seem comfortable being interviewed by AI.
A survey of roughly 300 respondents across the US and UK found that 73% of UK respondents and 76% of US respondents would regularly participate in AI-moderated research. Only around 10% said they were unlikely to do so.
For consumer insights teams, this removes a major concern. The problem is not whether people will speak to an AI interviewer. Most will.
But will they be honest?
Around half of the respondents said sharing their opinions with an AI interviewer felt as authentic and reliable as speaking with a human researcher.
This does not mean AI is as good as an experienced moderator. However, the absence of a human interviewer may not affect response quality as much as we assume.
In fact, AI could work better for certain topics. People often edit themselves when another person is listening, especially when discussing money, health, relationships, or behavior they feel embarrassed about. An AI interviewer that doesn’t judge them, lose patience, or react visibly may get answers that a human interviewer doesn’t.
The biggest finding, though, is about voice.
58% of UK respondents and 62% of US respondents said they could provide more detailed answers through audio or video than through writing.
This is important because consumers don’t think in neat, survey-friendly sentences. They pause, contradict themselves, remember something mid-answer, and use language the research team may never have considered.
A text box kills much of this detail.
Voice allows people to explain themselves naturally. The AI interviewer can then ask why, follow up on an interesting comment, or ask for a specific example. This brings some depth of a human interview into research that can be conducted with hundreds or thousands of consumers.
That is the real opportunity. Not another faster survey tool, but the ability to ask follow-up questions at scale.
There is one clear warning: don’t force people to turn on their cameras.
Only 37% of respondents in the UK were comfortable recording audio or video, compared with 49% in the US. In fact, 48% of UK respondents were uncomfortable with it (via Cint research report).
Audio, video, and text should be choices, not requirements.
Respondents need to know how their recordings will be stored and used.
For consumer insights leaders, AI-moderated interviews are interesting not because they make existing research cheaper, but because they enable a different kind of research.
Instead of commissioning a new project every time the business has a question, teams can speak to consumers regularly, follow their answers in real time, and build a deeper understanding of why they buy, switch, or leave (via OPXP.AI).
Consumers appear ready for this. The question is whether research teams will use AI merely to automate surveys or rethink how frequently and deeply they can listen to customers.