Lionbridge: 25 Years of Innovation in Localization
Machine Translation technology
Machine Translation (MT) has been around since the 1950s but took decades to progress to a point where the industry used it broadly. The simplistic rules-based approach made way for improved statistical models, which led to more advanced neural models and better—though not perfect—quality.
“The doom-and-gloom prediction was the most over-the-top,” Casal said. I remember in 2013-2014 as the first neural engines came out on the market people said, ‘The language industry is dead. It will be gone in five years.’ Here we are, stronger than ever.”
LSPs, like Lionbridge, recast themselves as post-edit companies. While MT machines might be able to effectively handle 80 percent of the load of a particular project, humans would still be needed to handle the remaining 20 percent of the material to get it to a publishable quality level.
Lionbridge was an enthusiastic adopter of MT and viewed it as a supremely powerful tool that required understanding, respect and confidence. Again, Lionbridge’s strategy was to boldly move forward, embrace MT and become the expert on the technology.
Lionbridge learned when to use machines, when to use people, which MT engine to use, when training is appropriate and cost effective, and what workflow is best. See our assessment tool in action on our MT Translation Tracker webpage.
Lionbridge obtained expertise in:
Lionbridge also developed real-time translation technology for unattended MT that is more secure than offerings like Google Translate.
MT enables Lionbridge customers to localize content that would otherwise be cost-prohibitive. As a result, they are able to localize more content.
Artificial Intelligence technology
The next major advancement, and the one that Lionbridge is currently addressing, is Artificial Intelligence (AI) technology. Like the three previous waves, AI portends to be a threat to any LSP’s survival if the LSP fails to adopt the technology or it doesn’t have enough data to get a sufficient confidence level in its machine learning algorithms.
Lionbridge is currently using its AI work to increasingly automate the creation and optimization of workflows. For instance, instead of humans making decisions about which translator to use, the idea is to use the body of data Lionbridge has already amassed about translators—with respect to their on-time delivery records, service quality and Language Quality Inspection (LQI) results—to pinpoint the best translator for the job.
Lionbridge has already started using applied AI to select translators and is in the process of developing other ways the technology can be used to automate localization processes, such as executing file prep or leveraging a particular set of content against a particular TM.
Furthermore, Lionbridge uses its big data to understand global content performance and global content effectiveness. For instance, within India, Lionbridge knows how a Tamil user reacts differently from a Hindi-speaking user. Having this knowledge enables Lionbridge to be their customers’ global partner in navigating global conversations.
“Every piece of content exists for a reason, whether it supports a marketing campaign, click or regulatory need. How well we help our customers support those needs globally is the next frontier. And how we use machine learning to achieve that and leverage all the data that we already have—not just for linguistic purposes, but also for content performance and workflow optimization purposes—that’s the next frontier,” Casal says.
What Other Challenges Did Lionbridge Overcome?
In addition to technology challenges, Lionbridge—like all companies operating in the beginning of the 21st century—would confront the dotcom bubble burst in 2000, the Global Financial Crisis of 2007–2008 and the COVID-19 pandemic, which continues to present societal challenges.
Lionbridge responded by looking for ways to further strengthen existing partner relationships. Lionbridge offered creative, cost-saving measures to help customers get through hardships. In the case of the coronavirus pandemic, Lionbridge quickly moved to a remote model without interruption to operations as lockdowns further underscored the need for digital transformation and global e-commerce.
Lionbridge has encountered many obstacles during the past 25 years. It has met each challenge and proven to be steady and secure.
How Did Lionbridge’s Acquisition Strategy Strengthen the Company?
Lionbridge acquired numerous companies over the years to enhance its services and technical capabilities. The following acquisitions are among its most notable:
Mark Kelly, a Dublin-based account director who joined Lionbridge as part of the Bowne Global Solutions acquisition, described the 2005 deal as a way for Lionbridge to demonstrate its strength in depth to its larger customers.
“Some of the customers we deal with are multibillion-dollar industries. They need to understand that they are working with strong, stable partners,” Kelly says. “When they see a very big player with a global presence working with all these international organizations, it gives them confidence to be working with a company like that.”
The Lions: How Have Global Misfits Been Critical for Success?
Cowan credits a lot of Lionbridge’s success to its culture and the intentional assembly of what he reverently refers to as “global misfits”. They’re Lions who embrace contradictions. An American who speaks Chinese? Yes, that’s a global misfit. So is a German who has studied anthropology. But it’s more than that, according to Cowan.