The Cables Are Too Ugly: Conversational Research and the Hidden Friction of Culture

The French-born psychiatrist and cultural anthropologist, G. Clotaire Rapaille, spent decades developing what he calls "archetype discovery," a method he adapted from psychoanalysis to uncover the emotional codes hidden inside a culture's relationship to a product, a brand, or an idea.

One of his best-known stories comes from his work advising AT&T. The American telecommunications company had a Japanese client, NTT, that sent over a detailed list of technical specifications for an order of cables. AT&T verified, point by point, that the shipment met every single requirement. Yet when the shipment arrived in Japan, NTT rejected the order outright; the cables were simply too ugly. AT&T’s engineers were astonished. The cables were destined for underground burial, far out of sight, and aesthetic appeal had never appeared on the specification sheet.

But for NTT, aesthetics was not a footnote to quality; it was, in Rapaille's reading of Japanese culture, an archetype of quality itself, inseparable from craft and, ultimately, from soul. The ugliness of the cables told NTT something about how little care AT&T had put into the work, no matter how many boxes had been checked.

I keep returning to that story because it captures something market researchers rediscover every time we move into a new region: meeting the brief is not the same as meeting the culture. A question can hit every specification on paper (correct language in a context) and still be rejected, quietly, by the people answering it. They won't tell you it's ugly. They'll just disengage, answer politely and vaguely, or drop out halfway through your research.

From Archetypes to Cultural Frameworks

Erin Meyer, professor at INSEAD and author of The Culture Map, built a structured tool for a similar underlying challenge. Drawing on research across more than thirty countries, she maps cultures along eight practical scales (i.e., how people communicate, evaluate, persuade, lead, decide, trust, disagree, and manage time). While Rapaille's work operates at the level of the unconscious (the emotional imprint formed early in life that a culture attaches to products, from coffee to cars), Meyer's framework focuses on how a culture behaves in the room; it clarifies what elements like politeness, silence, or a delayed reply actually mean. This system proves particularly useful when designing multi-country research projects. Meyer's scales provide a precise methodology to interpret and contextualize a participant's answer relative to a neighboring market.

Bilendi's latest ebook, How to Conduct Online Research in Latin America, applies Meyer's framework to examine six key markets: Argentina, Brazil, Chile, Colombia, Mexico, and Peru. This approach challenges a common industry assumption that sound methodology translates unchanged across borders. Global research often treats Latin America as a single homogenous block; it is broadly categorized as Spanish-speaking, mobile-first, young, and digitally engaged. While these general traits are directionally true, relying solely on them overlooks critical regional differences (e.g., Portuguese-speaking Brazil) and risks fundamental cultural misunderstandings: 

• Peru shows the strongest tilt toward high-context communication, where meaning is implied rather than stated outright. A question that reads as neutral in a US or European survey can land as abrupt in Lima.

• Mexico and Peru show the most pronounced hierarchical orientation in the region. In a focus group, that means a senior or dominant voice can quietly suppress the rest of the room, and it's on the moderator, not the respondents, to notice. Argentina is the outlier here, comfortable giving direct negative feedback in a style closer to Northern Europe than to its neighbors. Argentina prefers principles first, building the theoretical frame before the data point; Brazil, Peru, and Mexico lean applications-first, wanting the conclusion before the argument behind it.

• All six markets studied lean relationship-based rather than task-based; trust isn't earned through a well-run session; it's built beforehand, through small talk and genuine interest in the person answering.  The region also skews flexible-time; a respondent arriving late hasn't disrespected the schedule; they've simply prioritized something the schedule didn't account for.

AT&T's engineers weren't wrong; they delivered exactly what was asked of them. What they missed is that "meeting the brief" is a culturally specific idea, and specifications are only ever a partial description of what a market actually wants. A survey can be demographically balanced, translated correctly, and fielded on time, and still misread the room, because balance and translation were never the whole brief.

The deeper implication for global and multi-country research is that cultural adaptation isn't a layer you add after the methodology is set; it's part of the methodology. Deciding who counts as the household decision-maker in Mexico, how much time to build into fieldwork in Peru, or whether a focus group in Buenos Aires can tolerate open disagreement (it usually can, more than most) are not soft considerations. They determine whether the data you collect reflects what people think or simply what they were willing to say out loud to a stranger on a schedule that wasn't theirs. 

When the Survey Starts Talking Back

For most of the market research industry's history, cultural dimensions have mattered more in qualitative work. A focus group moderator, a doorstep interviewer, or an ethnographer reading a room could naturally adapt to local norms in real time. Quantitative research has never been fully insulated from this, and most researchers know it: differences in how cultures use rating scales, for instance, are well documented, and multi-country studies routinely correct for them. What is changing is not whether quant is exposed to cultural dynamics, but how directly.

Conversational surveys, AI-moderated interviews, and chat-based research tools now ask open, adaptive questions at a quantitative scale; they probe, rephrase, and follow up much like a human moderator. Because the instrument has become a conversational partner, it inherits the full range of dynamics described by cultural frameworks: how directly to request criticism, how much hierarchy or authority a respondent perceives in the tone, how long to tolerate silence, or how much rapport-building is needed before someone opens up. 

This is precisely where Rapaille's and Meyer's approaches need to work together rather than separately. Archetypes tell us what is emotionally at stake for a respondent in a given culture; Meyer's scales tell us how that respondent is likely to communicate it, or withhold it, inside a live exchange. A conversational tool designed without either risks the AT&T mistake at scale: technically flawless, fielded correctly, and still quietly rejected by the people it was built for, this time across millions of automated conversations instead of one shipment of cables. 

The Takeaway for Researchers

Consumers, much like NTT, rarely explain their rejection in terms a spec sheet can capture. They simply walk away; in research terms, they abandon, satisfice, or give whichever answer ends the conversation fastest. The lesson from Rapaille’s cables, Meyer’s scales, and Bilendi’s regional data points to a singular conclusion: cultural fluency is not a layer applied on top of good research design.

As our tools become more conversational and the line between qualitative nuance and quantitative scale blurs, cultural fluency becomes the research design itself: 

• Audit Current Instruments: Review existing international research tools and scripts against established cultural frameworks (e.g., Meyer's Culture Map or Hofstede's dimensions). Avoid over-simplifying complex regional nuances into rigid, one-size-fits-all rules.

• Redesign and Train Conversational Protocols: Build intentional rapport-building phases and contextual flexibility into AI-driven and live interview scripts. Monitor dropout rates carefully; over-engineered or artificial warm-ups can drive abandonment up rather than down. 

• Establish Cultural Quality Gates: Implement a mandatory cultural review before deploying quantitative tools globally. Measure success using concrete indicators (e.g., increases in open-ended response length, data richness, and completed interview depth).