By Emily Bary
Illustration by Glenn Harvey
Money talks, and soon it will be easier to understand.
Language barriers can hinder international commerce, but that problem is increasingly being solved by machine-based translation programs, which are becoming sophisticated enough to enable people in different countries to communicate as if they spoke the same language.
Lily Chen, a sales manager with electronic-forklift company Taixing Jichuan Hydraulic Machinery Co. Ltd. in China, said she’s used Alibaba Group Holding Ltd.’s /zigman2/quotes/201948298/composite BABA -0.15% built-in translation tools to communicate with buyers in Europe and the U.S. In an interview with MarketWatch using the technology, she explained that her English was “not very good” when she began as a salesperson and the translation software first helped her conduct professional communications with a customer in Germany and, later, with other buyers.
“I was a little nervous at first, but from the customer’s reply, the customer fully understood the meaning of the translation, which made me feel very confident,” Chen said. She deemed the software able to handle “the professional vocabulary of electric forklifts,” including fork width, loading capacity and after-sale service.
Translation programs aren’t yet tracked specifically by economists, but they have the potential to boost international trade, which stood at $19.5 trillion last year according to World Trade Organization estimates.
E-commerce’s international expansion
One big opportunity exists in e-commerce, where the giants of online shopping are already beginning to incorporate translation technology into their businesses. Take eBay Inc. /zigman2/quotes/204653455/composite EBAY +0.44% , the San Jose, Calif.-based online marketplace, which quickly learned that while the internet has made it easy to connect buyers and sellers, international expansion wouldn’t be particularly effective if shoppers weren’t able to understand product listings from abroad.
The result is a system that enables shoppers to make search queries in their preferred language and receive translated product listings, while also taking into account the context of a customer’s request. Machine translation has to be smart enough to adjust its behavior based on whether someone is looking for a Galaxy Note 10 smartphone or a galaxy-print sweatshirt, for example, since the branded product doesn’t need to be rendered into a different language.
“Anything that supports cross-border trade so that someone from outside the U.S. is capable of engaging in e-commerce is where the power of global trade lives,” said Sanjeev Katariya, an eBay vice president who focuses on artificial intelligence.
EBay’s push for machine translation has helped the company increase Latin American exports by nearly 20%, according to researchers from the Massachusetts Institute of Technology, and illustrates the potential for increased commercial activity as translation technologies gain wider adoption in business.
“By embedding translation tools, you’re fostering the most critical thing in trade: trust.”
John Caplan of Alibaba
The positive impact from translation ties into the “gravity” theory of trade, explains MIT professor Erik Brynjolfsson, in that trade between countries is usually dependent on how close they are geographically. “You can reduce transportation costs and make the world smaller by 20% or you can introduce translation and have the same effect,” he said in an interview.
Machine translation vs human translation
The stakes are high for retailers as e-commerce gets more global. In order to be accessible in the places where 90% of the world’s online spending happens, businesses must offer support on their sites for the 15 most economically beneficial languages, according to Donald DePalma, the chief research officer at Cambridge, Mass.-based CSA Research. Just four years ago, they would have only needed to feature 11 languages. To reach even more spenders -- 99% of the online marketplace -- today requires support for 56 languages.
Yet many sites offer support for far fewer. CSA looked at some 2,800 of the most trafficked sites globally and found that only 63% were multilingual. Those featured support for 7.8 languages, on average, in addition to the main language.