What Synthetic Consumers Cannot Hear

Synthetic consumers have entered the language of contemporary research with a powerful promise: scale, speed and the ability to simulate future scenarios before they fully emerge in the market. Used carefully, they can support strategic thinking, test hypotheses and expand the analytical imagination. Yet their growing presence also raises a deeper methodological question for qualitative research. What happens to the authority of the qualitative researcher when simulated profiles begin to speak as if they were real people? This article argues that synthetic consumers should not be rejected, but their limits must be clearly understood. Simulation can help us explore possibilities, but it cannot replace listening, presence or the human capacity to recognise cultural tensions before they become stable patterns in data.

There is something quietly seductive about the idea of synthetic consumers. They promise scale, speed and the simulation of complex scenarios in just a few clicks. In an organisational culture that values predictability and rapid decision-making, the proposition can feel almost irresistible.

If we can model patterns, why not model people? If we can project behaviours, why not anticipate trends? The appeal is clear, especially in contexts where uncertainty feels costly and time for interpretation seems increasingly scarce.

But this is exactly where my unease begins.

When qualitative research starts to rely on synthetic consumers to study trends, it shifts something fundamental in its own centre of authority. Traditionally, the authority of the qualitative researcher comes from listening. It comes from being present in in-depth interviews, focus groups and ethnographic encounters.

It comes from the ability to stay with ambiguity when a participant hesitates before answering, and from the discomfort of dealing with what has not yet become organised or coherent. The qualitative researcher is not only someone who collects statements. The qualitative researcher is someone who can recognise cultural tensions while they are still taking shape.

Synthetic consumers, by definition, are constructions based on data that has already been captured. They often combine quantitative datasets, previously coded qualitative material and narrative patterns converted into variables. They are structured projections of profiles, discourses and likely reactions.

Even when they are sophisticated, they remain reconfigurations of what has already been seen, measured or interpreted. They operate through extrapolation. That does not make them useless, but it does define the boundaries of what they can know.

In a recent meeting, synthetic personas were already being presented almost as if they were real participants. There were detailed descriptions of their motivations, simulated responses to new stimuli and projected consumer journeys with apparent internal coherence.

The exercise was sophisticated. The narratives sounded organic. They were easy to follow, easy to discuss and easy to believe.

Then the client, after carefully observing the simulations, asked a simple question. Where, in this model, is the possible market tension around the international tariffs that have already been announced and are expected to increase further?

There was a pause.

The personas had been built from consolidated data and previously identified patterns. But that geopolitical tension had not yet become an analytical category. It was still in the field of uncertainty, diffuse concern and conversations crossed by unease. It had not stabilised as a variable.

For that reason, it did not appear in the simulations. Yet it was already affecting decisions.

This is where excessive confidence in models begins to reveal its limits. A trend is not always an extrapolation. Perhaps I am being cautious, but many transformations begin precisely in the territory of what has not yet gained methodological consistency.

In interviews, this may appear as an interrupted answer. In a focus group, it may emerge as a side comment that no one develops. In ethnography, it may appear as a gesture that does not quite match the declared discourse. Not as a consolidated pattern, but as a weak signal with cultural weight.

When organisations begin to rely almost exclusively on synthetic consumers to anticipate the future, something delicate happens to the authority of the qualitative researcher. The researcher risks moving from interpreter of living complexity to validator of previously modelled scenarios.

Listening, which once opened the path to understanding, can become a stage of confirmation. The field no longer begins the inquiry. It is invited to approve what the model has already made plausible.

There is a silent inversion here.

If the researcher’s authority becomes based mainly on the ability to operate models and translate simulations, it moves closer to technological authority. But qualitative authority is of another nature. It is born from interpretation, presence and the capacity to perceive what has not yet organised itself into a coherent narrative. 

The greatest risk is not the use of synthetic consumers in itself. They can be useful for challenging hypotheses, exploring possibilities and accelerating strategic reasoning. They may help teams imagine scenarios that would otherwise remain abstract.

The risk lies in the illusion of completeness. It lies in the feeling that, because simulated profiles respond consistently, the scenario has been sufficiently understood. Sometimes it has not. More often than we may want to admit, it probably has not.

At times, the model organises the past with such clarity that it softens the ruptures beginning to emerge. By transforming previous narratives into predictable patterns, it can make invisible the tensions that have not yet been clearly verbalised.

And in doing so, it produces comfort.

For the qualitative researcher, this creates greater responsibility, not less. If technology occupies more space in the construction of scenarios, human authority must move to an even more reflective place. It is no longer enough to know how to operate tools.

We must know how to question them. We must ask what was left outside the model. We must recognise when the context is changing faster than the available data can capture.

In this context, the researcher’s authority no longer lies in possessing the answer. It lies in the courage of the question.

Perhaps the future of qualitative research is not about resisting synthetic consumers, but about defining their limits clearly. It is about affirming that simulation is not listening. Projection is not presence. Extrapolation does not replace cultural sensitivity.

Because there are economic, social and symbolic movements that begin before any database can register them. They emerge in interrupted sentences, hesitant looks and tensions that are still poorly formulated. They only become visible when someone is willing to listen before organising.

The question is not whether we will use synthetic consumers to study trends. We will.

The question is who will still have the authority to say: there is something here that the model has not yet seen. And whether we will have the courage to sustain that position when the simulations appear complete.