Beyond words: Multilingual customer service means understanding culture, not just language

I always wanted to go to Japan. When I was a child, I liked the Teenage Mutant Ninja Turtles. Which led me to ninjas, which led me to Samurai, which led me to the history of Japan. 

Then, during college, I had a chance to apply for a seven-week intensive study abroad program taking five students to Japan. Naturally, I leapt at the opportunity.

I had completed exactly one year of Japanese language at that point, and my recognition of Hiragana and Katakana was B- at best. My spoken abilities were basically non-existent.

I remember that first day in Japan, awake at 3am and giddy about being there, we found a convenience store on the first floor of our building. I couldn’t read a thing, but luckily clear packaging means I could tell a rice ball from… not a rice ball. So we grabbed a couple and took them to the counter.

The clerk rang up the items and then spoke what might as well have been gibberish. My whole year of study amounted to bupkis because there was no way I could tell the beginning from the end of whatever she said to me.

It’s 25 years later and I’ve passed my JLPT N2 exam. So now I can tell you, confidently, that she said: あたためますか or “would you like to heat it up”. To which I should have replied: はい – “yes”. I also know now that I probably shouldn’t have torn into the packaging with my teeth like a gorilla when there were perfectly clear instructions on how to open it.

I’ll never forget that encounter. I learned throughout those seven weeks that convenience-store rice balls are perfectly edible cold. But more importantly, I learned that I could barely survive in Japan without speaking the language.

Today, being a fluent Japanese speaker, I know the words to buy and heat up a delicious convenience-store rice ball. But the opportunities of language go much deeper than that. 

My friendships and professional relationships in the country are built on more than just the words we say to each other. They’re built upon an understanding of culture and etiquette.

Which is an increasingly important thing to consider today in a world of easy language translation. Organizations of any size can use AI or even simpler tools to talk to their customers in almost any language. But to truly connect with them? That takes an understanding of something deeper.

Multilingual customer service matters

First, let me say that multilingual customer service is no longer a luxury – it’s an expectation. Businesses aiming for global reach know that language barriers can create friction, frustrate customers, and even lead to lost revenue. The converse is also true.

A huge 76% of people say they’re more likely to buy products that are presented in their first language. Similarly, 74% will repeat business if support is offered in their language.

Almost two-thirds say they “rarely or never” buy from English-only websites.

Score one for Google Translate. But multilingual service is only step one in crafting great experiences. Building a deeper, personal connection with customers goes beyond words and into the realm of cultural understanding.

A customer in Tokyo might expect a formal, respectful tone, while a customer in New York might prefer a casual and direct approach. A simple “yes” or “no” may carry different connotations across cultures, and humor can easily get lost in translation. 

Even how people expect to be helped varies by region – some cultures expect proactive assistance, while others prefer a self-service model.

The hidden gaps in traditional multilingual support

Most multilingual customer service solutions rely on either:

1. Human agents who speak multiple languages but may not always be available or cost-effective.
2. Chatbots or AI translators that provide direct word-for-word translation but lack cultural sensitivity.

The problem with the first approach is scalability: hiring native-speaking agents for every language a company serves is expensive and difficult. The issue with the second approach is context: automated translations often fail to convey the right intent, tone, or cultural meaning.

For example:

• In Germany, customers expect precise, information-heavy responses with minimal small talk.
• In Brazil, a warm, conversational approach is more appreciated.
• In Japan, politeness and hierarchical respect are crucial – direct refusals may be considered rude.
• In the United States, a customer may expect problem-solving to be direct and efficient, with an emphasis on personal connection.

Traditional AI chatbots don’t account for these subtle differences. A literal translation might turn a polite Japanese response into something that sounds cold in English – or vice versa. This mismatch can lead to misinterpretation, frustration, and ultimately, poor customer experiences and negative brand association.

How digital humans bridge the cultural gap

Digital humans are designed to do more than translate. Driven by LLMs and/or NLP, they can offer multilingual customer service at scale like any AI interface. But the connection runs deeper.

Digital humans are designed to represent the cultures and communities in which they live. This is why we’re a huge proponent of workshopping the AI’s personality, backstory, quirks, and characteristics during the digital human design process.

Take Emma for the City of Amarillo, for instance. It’s Emma’s job to help residents find information on the City’s website using live, direct conversation. 24% of Amarillo residents speak a language other than English at home, and Emma is equipped to help them all.

However, the key to Emma is her personality. She was designed with the local community to be a strong representation of Amarillo’s people. 

She has a family history, a connection with the region, and her own goals and ambitions. It makes speaking with her a much more rounded experience for residents. They see themselves in her, making her more than just a lifeless chatbot.

Going beyond words: What else makes these AI experiences more effective?

So, new real-time AI avatar technologies can look the part. And they can speak 100 languages with ease. But there are other levels that make interactions even more engaging to diverse audiences.

• Cultural adaptation: Digital humans don’t just switch languages; they adjust tone, formality, and response style based on the customer’s cultural expectations. Research from the Journal of Cross-Cultural Psychology shows that communication styles vary significantly between cultures. Using prompt engineering, AI interfaces can easily reflect these nuances to feel more natural and personalized.
• Emotionally intelligent conversations: Digital humans can detect sentiment in real-time and adjust their responses accordingly. Frustrated customers can be met with a more reassuring, calming tone, just like a well-trained human agent.
• Non-verbal communication: In many cultures, body language, facial expressions, and even pauses in speech play a huge role in communication. It’s why Yuria, our digital human for the Japanese market, greets people with a formal bow.

The future of multilingual customer service

It sounds corny to say, but this is truly just the start of AI’s foray into better serving different cultures, providing experiences that are not just personalized but individualized. And it clearly goes far beyond the words that are said in conversation.

Companies like Meta and Google have already been investing in AI-driven language models that improve contextual translations. Meanwhile, researchers from MIT and Stanford have been exploring how AI can detect and adapt to cross-cultural sentiment analysis to make interactions more human-like.

At the start of this year, we were delighted at UneeQ to help Pearson Languages launch its AI language tutor to help with corporate language training. Things are heating up as more organizations explore the power of AI for language applications.

While human customer support is the gold standard, many businesses with diverse customer bases will see tremendous gaps in how effective their people are at customer care and how poor their digital solutions are.

AI will close that gap. But only if solutions are designed for humans – to meet our human need to feel supported, understood, and cared for. And that’s true in any language.

‍