From the Why to the Impact

Long conversations instead of quick clicks — we listen until behaviour makes sense.

Scroll

Stance

People are not data points.

The industry takes refuge in dashboards. We sit down and listen: in long, depth-psychological conversations in which people reach out, digress, contradict themselves — and in doing so reveal why they do what they do. No scripted questions, no scripted answers, no bots.

“When the data and the anecdotes disagree, the anecdotes are usually right.”
Jeff Bezos — the sentence behind it: then something is wrong with the measurement.

Bots can’t sit on the couch

Online panels are filling up with scripts and language models. A two-hour face-to-face conversation is the one research format that cannot be botted.

Numbers say where. Conversations say where to.

Metrics describe the present of what already exists. Decisions are about what doesn’t exist yet — and that never shows up in a dashboard. A questionnaire only measures its author’s imagination.

The anecdote is the early-warning system

Trend breaks first appear as a single person saying something strange in an interview — months before they become significant. Once the curve moves, it is news, not insight.

Fig. — Dots alone say nothing. Only in conversation do they condense into the question that explains behaviour.

We are not against numbers. Quantitative data rounds off our studies — we just don’t hide behind it.

Method

We research where others don’t even look.

Since 1986 we have used morphological depth psychology to decode the unconscious patterns of meaning and structure behind behaviour and experience — in the rheingold salon Interview®, in Groups®, in workshops. Theory-led instead of gut feeling, independent of any agency network.

Fig. — The course of an analysis: from exactly one angle, what was said arranges itself into a gestalt.
Get to know the method

Impact

The best bird food was too good.

Living-room interviews revealed that some bird owners regretted buying their bird — and resolved the conflict, unconsciously, in the food aisle, with the second-best food. Answers like these appear in no dashboard. Four case stories show what becomes of them.

Read the case stories

AI in the service of depth

Machines transcribe. People answer.

Language models are photographs of yesterday — people experience new things and change their minds. That is why no synthetic “respondents” replace the conversation here. Everywhere else, AI has long been working for depth: co-creation with image models, transcription in dozens of languages, agent-assisted analysis on our own infrastructure.

Even the big AI labs filter machine-generated text out of their training data and invest in watermarking to keep it out. Whoever acts that way treats human data as what it is: irreplaceable.

Fig. — A model extends yesterday. People turn away — and exactly there the signal appears.
Our AI tools in detail

Salon

A place where thoughts gain depth.

The salon is how we work: interviews, analysis sessions, events, eating and drinking together — in the heart of Cologne, right by the cathedral.

Get to know the salon
Jens Lönneker, full figure Ines Imdahl, full figure Lukas Struwe, full figure
Fig. — Jens Lönneker · Ines Imdahl · Lukas Struwe

News

Publications & events

KI · Methode

The Turing test proves nothing about markets

That a language model is indistinguishable from a human in conversation tells you nothing about whether it tracks real market trends. A perfect actor is not a witness.

rheingold salon
Studie · Kultur

The 50plus generation: invisible to dashboards

Over-50s hold the majority of purchasing power and barely register in click metrics.

Ines Imdahl
Studie · Medien

Trust in media: a deep core sample

What people call loss of trust is rarely distrust of facts — it is the feeling of no longer appearing in one's own everyday life.

Jens Lönneker
  • Podcast: Warum Menschen kaufen, was sie nicht brauchen podcast · online
  • Salongespräch: Das Warum hinter den Zahlen event · rheingold salon, Köln

Digital independence

Our research runs on our own infrastructure. Interviews, transcripts and analyses never leave our servers.

This website loads nothing from third parties and sets no cookies — which is why it never asks about them.