Your Design Doesn't Translate

Every design is a cultural document. You don’t choose to make it one. It just is, because every assumption baked into how your product works (who it expects users to trust, how it thinks decisions get made, what it treats as obvious) comes from the culture you grew up in. You can’t design your way out of that any more than you can unlearn your first language. What you can do is notice it.

Most designers don’t. They ship into new markets and wonder why something that works in one place falls flat in another. They translate the copy. They adjust the currency. They assume the rest is decoration. It isn’t.

Every design carries invisible logic. Not in the code, not in the copy, but in what the design takes for granted about how people make decisions, who they trust, and what has to happen before money changes hands. That logic is cultural. You built it without knowing you were building it, because it’s the water you swim in. And when your design lands somewhere that runs on different water, it doesn’t translate. It breaks at the seams nobody thought to check.

The programming nobody sees

In the 1970s, Dutch social psychologist Geert Hofstede spent a decade analyzing survey data from IBM employees across more than 70 countries. What he found wasn’t that people in different places had different opinions. He found that they had different mental architectures. Different assumptions about authority, uncertainty, relationships, and trust, all of them learned in childhood, invisible to the person holding them, and resistant to change.

He called it “the collective programming of the mind that distinguishes the members of one group or category of people from others.” That phrase does more work than it looks like. Programming, not preference. Collective, not individual. The implication is that your users aren’t making a conscious choice to trust or distrust your interface. Their reaction is faster and older than that.

One of Hofstede’s key dimensions maps straight to how interfaces either earn trust or fail to. He called it individualism versus collectivism. High-individualism cultures, which include the United States, the UK, and most of Northern Europe, expect individuals to operate as sovereign agents. You see a product, you read the listing, you check the rating, you buy. The interaction is between you and the object. Trust is encoded in the system. Low-individualism, collectivist cultures, which include most of East Asia, Latin America, South Asia, and Sub-Saharan Africa, don’t work that way. Trust isn’t in the system. It’s between people. Before you buy, you need to know something about the seller. You want to talk. You want signals that the person behind the listing is real and accountable in the way your culture understands accountability.

Edward Hall mapped a related dimension in 1976. He distinguished between low-context cultures, where communication is explicit, direct, and self-contained, and high-context cultures, where communication relies on shared background, implicit signals, and relational context to fill in what the text doesn’t say. A high-context user finds it cold, thin, and blunt in a way that reads as dismissive.

This isn’t about aesthetics. It’s about what triggers trust and what blocks it. Designers who don’t know which culture they’re building for build for themselves by default.

Three hundred million dollars and a missing chat window

In 2002, eBay entered China by acquiring EachNet, a local startup that held about 85% of China’s online consumer-to-consumer market, paying $180 million for full control by 2003. They sent in their own management. The general manager was German. The CTO was American. Neither spoke Chinese. They migrated EachNet’s platform to eBay’s American infrastructure, a move that cut site traffic in half on the day it went live.

eBay’s Chinese users found clean listings, structured ratings, fixed transaction fees, and PayPal. What they didn’t find was a way to talk to the seller before buying. eBay’s interface assumed a ratings system was sufficient evidence of trustworthiness. That assumption worked in the American market. In China, it was a wall.

In May 2003, Alibaba founder Jack Ma launched Taobao as a direct competitor. Taobao was free. It had AliWangwang, a built-in instant messaging tool that let buyers and sellers talk before any money moved. It had Alipay, an escrow system holding payment until the buyer confirmed receipt. Its colors were red and orange. Its design was dense and loud, packed with links and promotions, reflecting the sensory environment of Chinese commercial life. The name translates as “digging for treasure.”

What Ma understood, and eBay didn’t, was guanxi. It’s a Chinese concept that describes the web of close personal relationships, mutual obligations, and earned trust that underpins how commerce works in Chinese culture. In a context where legal dispute mechanisms were weak and online fraud was a real concern, buyers needed to build something with a seller before handing over money. Not a star rating from a stranger. A conversation. A sense of the person. Taobao built that channel into the product. eBay’s design made it impossible by design.

The market spoke without ambiguity. Between 2003 and 2005, Taobao’s market share rose from 8% to 59%. eBay’s fell from 79% to 36%, then to 29% by 2006. eBay shut down its Chinese operations that year after spending close to $300 million. By 2008, Taobao held over 80% of the market. Jack Ma described it this way: “eBay may be a shark in the ocean, but I am a crocodile in the Yangtze River. If we fight in the ocean, we lose, but if we fight in the river, we win.”

eBay had a functioning product. It had brand recognition, capital, and first-mover advantage. It lost to a smaller company that understood whose trust it needed to earn and what that trust required. The design was not wrong. It was wrong for the cultural context it operated in.

One question before you ship

The mistake isn’t failing to hire a translator. Most teams do that. The mistake is treating translation as the whole job, when translation only handles the surface. The deeper work is asking whose mental model your design was built for, and whether the people you’re shipping to share it.

There’s a simple test worth running before any cross-cultural launch, or any expansion into a user base different from the one you built for. Ask this about your core transaction flow: does this design assume the user already trusts a stranger?

If your flow relies on a ratings system, a review count, or any mechanism where trust is delegated to strangers rather than earned through direct interaction, you’re building for a low-context, high-individualism user. That works well in some markets. In others, it’s the exact thing that breaks trust rather than building it.

The fix isn’t always a messaging system. Sometimes it’s visible founder identity. Sometimes it’s community signals showing the seller sits inside a real network, not floating as a stranger in a marketplace. Sometimes it’s a face and a name that feel real in the culture you’re entering. The form changes. The question stays the same.

Bring someone from the target culture in before design, not for a translation review at the end. Ask them to describe how commerce works in their life. Where does trust come from? What signals it? What would make them stop? Build those signals into the design. That’s not localization. That’s designing for a real person instead of a projection of yourself.

The document your design already is

Every interface encodes assumptions about users. Who they trust, how they decide, what they need before they act. Those assumptions feel invisible to the person who made them, because they come from the culture the designer lives in. They only become visible when the design lands somewhere that culture doesn’t reach.

Your design isn’t a neutral object. It’s a cultural document. The question is whether the people you’re building for are the people it was written for.

⸻