By ICG member Frank-Thomas Naether

I have always been convinced that in a world that is becoming increasingly digital, qualitative market research will gain in importance. I remain convinced of this, especially in the age of AI.

The new possibilities offered using AI open the door to a completely new self-image of qualitative market research and the qualitative market researcher.

The role of moderation in the age of AI and an outlook

Introduction

The digital transformation has fundamentally changed the market research industry. While artificial intelligence is taking over more and more analysis processes and processing data in a matter of seconds that used to take days, an apparent paradox is emerging: the human factor is becoming increasingly important in qualitative market research in particular.

This shift in importance relates in particular to an aspect that is often overlooked but is crucial for the quality of the results: the moderation of group discussions and individual explorations.

My thesis: The quality of qualitative research results depends directly on the quality of the input – and this is largely determined by the expertise of the moderator. An experienced, qualified market researcher with strong interpersonal skills makes all the difference here.

Why is good moderation so important?

Qualitative market research is a profound process of understanding in which we try to grasp not only what people do or say, but also why they do or say it. In this context, the moderator is the central interface to deeper understanding on the one hand and to the general course of a conversation on the other.

There are essentially three areas in which the relevance of good moderation comes into play:

1. Emotional intelligence:
   • Empathy: The moderator can create an atmosphere of trust in which the participants feel comfortable and openly share their thoughts. This is particularly important for sensitive topics.
   • Non-verbal communication: The moderator can interpret and react to the body language and tone of voice of the participants in order to gain deeper insights.
2. Flexibility and adaptability:
   • Situational action: The moderator can react to unexpected turns in the discussion, adapt questions or take up new topics when it seems relevant.
   • Group dynamics: The facilitator can control the dynamics in the group to ensure that all participants have their say and that different perspectives are taken into account.
3. Interpretation and contextualisation:
   • Deeper understanding: The facilitator can interpret the participants’ responses in the context of their personal experiences and background.
   • Synthesis: The moderator can synthesise the various contributions in order to obtain a comprehensive picture of the topic.

In the following, I will go into more detail on individual aspects, focusing on the relevance of good moderation in principle as well as on the possibility, which has been discussed more frequently recently, of AI taking over moderation in the future (solutions are already being offered that go far beyond chatbots).

Empathic understanding and emotional intelligence

A human facilitator has the natural ability to recognise and understand emotional nuances. For example, if a participant responds hesitantly, the moderator can recognise whether this reluctance stems from uncertainty, personal discomfort or perhaps the need for more time to think. These subtle distinctions enable the moderator to respond appropriately – for example, by using reassuring words, rephrasing the question or deliberately pausing to create space for deeper reflection.

Dynamic adaptability

The course of a qualitative interview or focus group is rarely linear. An experienced human moderator can spontaneously adapt the course of the conversation when interesting or unexpected topics arise. This ability to improvise makes it possible to gain relevant insights that might remain undiscovered in a more rigidly structured format. The facilitator functions here like an explorer who follows new, promising paths as soon as they appear.

Building a basis of trust

Trust is the currency of qualitative research. People are more likely to open up and share authentic thoughts when they feel safe and understood. A human facilitator can build this trust through subtle signals – eye contact, affirming nods, respectful listening and showing genuine interest. This foundation of trust leads to more honest and deeper responses than in environments where this human element is missing. I think this aspect is absolutely key! A ‘non-human’ will never be able to do this in the same way as a professional moderator, because there is no script or template for this, it is new and different for every interview and every group or workshop.

Cultural context sensitivity

Human moderators have an intuitive understanding of cultural norms, taboos and communication styles. They can recognise when a topic is culturally sensitive and proceed with appropriate caution. This cultural intelligence is particularly valuable in globalised markets or when conducting research with diverse target groups, where cultural misunderstandings could significantly affect data quality.

The art of in-depth questioning

The true art of qualitative moderation lies in asking questions – recognising when a superficial answer deserves deeper exploration. A skilful moderator can gently but effectively probe: “Can you tell me more about that?” or “What led you to that opinion?” This technique of “peeling the onion” often leads to the most valuable insights that go well beyond the initial answers. Will AI ever be able to outweigh experience? Despite the amazing capabilities and developments of AI, I have my doubts.

Here is an example from practice: “In a group discussion on banking products, one participant initially answered the question about his preferences superficially: ‘I’ll just take the cheapest offer. Instead of accepting this statement, there was a noticeable discrepancy between the verbal statement and the participant’s sceptical posture. By asking specific questions – “What does favourable actually mean to you personally?” – a completely new area of insight opened up: the participant did not define ‘favourable’ by price, but by a balanced relationship between performance, trust and costs. This differentiation would have been lost without intuitively recognising the contradiction and asking empathetic questions.”

Balancing the group dynamics

In group environments such as focus groups, social dynamics inevitably arise. Some participants dominate the conversation while others are more reserved. A human moderator can pick up on this dynamic and balance it out – by tactfully limiting more dominant individuals and encouraging quieter participants. Without this deliberate control, the findings might be dominated by a few loud voices, which could lead to biased data collection. And anyone who regularly moderates knows the situation when a mood is in danger of tipping over from a certain point. A ‘real’ moderator can, no, must be able to react quickly and appropriately and correctly assess all participants in the discussion simultaneously and in real time.

Here is another practical example: in a group discussion on the topic of sustainability in food, one particularly committed and rhetorically skilful participant threatened to dominate the discussion. Other participants withdrew and reduced their contributions to nodding in agreement. With a combination of gentle limitation (‘Thank you for your perspective, I would also like to hear other voices’) and targeted activation of reticent participants by addressing them by name but unobtrusively, it was possible to balance the dynamic. The subsequent analysis showed that the initially reticent participants in particular contributed important nuances that were crucial for understanding the target group.

Interpreting non-verbal communication

A significant part of our communication is non-verbal. A certain facial expression, a laugh or smirk or a sudden change in posture – all of these can convey important information that complements or even contradicts verbal responses. Good (and human) facilitators can recognise these signals and respond appropriately, adding an extra dimension of ‘data gathering’ that goes far beyond the spoken word.

Ethical and situational judgement

Qualitative research can sometimes touch on sensitive or emotional issues. It takes concrete experience to recognise when a participant is showing discomfort and handle situations with empathy. The ability to maintain ethical boundaries and ensure the well-being of participants is not only morally important, but also contributes to the quality of the research by creating an environment where participants feel respected and heard. It’s simple: the interviewee must feel comfortable!

Human moderation in qualitative market research is therefore more than just a data collection method – it is almost a complex art form that combines empathy, adaptability and social intelligence to gain deeper insights than would be possible using standardised or automated methods alone.

Another, and in my opinion extremely exciting, excursus revolves around the question of synergies between ‘classic craftsmanship’ (moderation by an experienced moderator) and AI: How can AI be used to serve the moderator and lead to significant progress in the method?

The limits of human moderation and the potential of AI support

Even experienced moderators come up against limits that can be overcome with AI support:

• Unintentional bias: Human moderators inevitably bring their own experiences, preferences and biases to the table. These can subtly influence the formulation or interpretation of questions. AI systems can recognise this potential bias by analysing the text and point it out to the moderator.
• Cognitive load: After several hours of intensive moderation, attention naturally wanes. AI assistants can run in parallel and provide discrete stimuli when the depth of enquiry decreases.
• Processing capacity: Humans can only process a limited amount of information at the same time. In complex discussions with many participants, important aspects can be overlooked. AI can follow conversations in real time and point out overlooked topics or contradictions.
• Documentation challenge: It is almost impossible to be deeply present in the conversation and precisely document relevant aspects at the same time. AI-supported systems can do this in parallel and even make initial categorisations.

The ideal solution lies in complementary collaboration, where AI acts as a discrete co-moderator that augments human capabilities without interfering with authentic human interaction.

Evolution of methodology through AI integration

The combination of human moderation and AI support not only leads to more efficient processes, but also enables methodological innovations in qualitative market research:

• Adaptive interview guides: AI can identify key topics during the interview and dynamically suggest customised questions tailored to the participants’ specific interests and experiences.
• Multi-dimensional analysis in real time: While the moderator concentrates on the interpersonal interaction, AI can carry out parallel sentiment analyses, create semantic networks and identify thematic clusters – all of which can discreetly serve as input for the moderator.
• Hybrid research designs: The combination of AI-supported pre-analysis of large amounts of data and targeted qualitative in-depth exploration enables new research designs that combine quantitative breadth with qualitative depth.
• Continuous learning loops: AI systems can learn from each moderated session and help moderators refine their techniques by identifying patterns of successful interventions.

These methodological advancements significantly expand the toolkit of qualitative market research without compromising its fundamental value – the deep understanding of human motivations and behaviours.”

Let’s utilise the best of both worlds: human creativity, empathy and social intelligence paired with the analytical power, speed and scalability of AI.

Outlook

I see three specific development paths in the coming years:

1. The augmented moderator: special, discrete AI assistance systems that support moderators during the discussion – for example by providing real-time feedback on overlooked topics or suggestions for more in-depth questions that are displayed via augmented reality glasses or subtle signals.
2. Hybrid research teams: Instead of the traditional distribution of roles, research teams are being created in which human moderators and AI systems work together in a complementary way – for example, by AI pre-sorting large amounts of data and human moderators going into greater depth in a targeted manner.
3. New qualification profiles: Qualitative market researchers are developing new skills profiles that include not only traditional moderation skills but also the confident use of AI systems – a ‘digital qualitative researcher’ who embodies the best of both worlds.

The key to success does not lie in the question of “man or machine?”, but in the intelligent utilisation of the respective strengths. Qualitative market research is not at the end, but at the beginning of an exciting evolution that will further increase its potential to generate deep insights.”

However, this article does not end without an important final note.

Risks of excessive technologisation

Despite all the enthusiasm for the new possibilities, critical aspects must not be ignored:

Technological dependency: An excessive shift to AI-supported systems could lead to an erosion of basic moderation skills, similar to how GPS systems have weakened our inner sense of direction.

Loss of authenticity: Participants sense when a facilitator pays more attention to their technical system than to themselves. The authenticity of the interpersonal encounter, which is at the heart of any successful qualitative research, could suffer.

Privacy and trust: Real-time analysis of sensitive conversation content raises complex ethical and legal issues that must be carefully addressed in order not to jeopardise the trust of participants.

Technological determinism: There is a danger that we will begin to focus our research questions on what is technologically easy to measure, rather than being guided by genuine interest in knowledge.

A considered integration of AI into qualitative research therefore requires not only technical understanding, but also a heightened awareness of these potential pitfalls.”

 

Frank-Thomas Naether: MD and Owner of Naether Marktforschung GmbH in Hamburg/Germany

Frank-Thomas is in the business for more than 35 years and has conducted more than 6.000 group discussion plus numerous IDIs.