In this XTRF Summit #online 2021 session, Jaime Punishill and Kajetan Malinowski, Chief Marketing Officer and Product Director from Lionbridge respectively, talked about what the future of localization looks like.
This article is based on that session.
The past year has been a tumultuous one, with the COVID-19 pandemic upending daily life for most people and their employers. Everything—from the way we work, shop, and buy products—has changed. Many of these changes will be long-lasting, if not permanent.
We’ve identified three localization trends you should be aware of and suggest what to expect in the future. While there’s no doubt that we’ve been through a lot, Language Service Providers (LSPs) may come out stronger following the adversity of the past year.
Top 3 trends in localization
1. The rapid acceleration of CX and WX digitization
The Covid-19 pandemic unexpectedly accelerated digital transformation. Though many companies had made steps towards digitization before 2020, the pandemic exposed just how poorly managed many of these efforts had been. Companies were suddenly forced to shift their internal and external processes to accommodate a virtual world, and those who failed to digitize were unable to successfully pivot.
Internally, business found that the new digital-first universe affected WX, or Workforce Experience. Every employee touchpoint had to become digital, from eLearning and training to virtual collaboration. The pandemic affected both systems and processes.
Externally, all aspects affecting the Customer Experience (CX) had to rapidly shift to accommodate online interactions. Mandatory lockdowns meant offline retail channels were no longer an option. Both customers and businesses had to adapt accordingly. Companies were faced with having to make a huge move to a virtual existence overnight.
Retailers—many of whom only localized a fraction of their digital content prior to the pandemic—had to drastically increase localization efforts to accommodate the way consumers were being forced to evaluate the content and shop for goods. Companies had to move fast. Similarly, companies needed to quickly find ways to effectively communicate and train their employees in a virtual manner.
Even in a post-pandemic world, there is no going back as customer habits and workplace changes have taken root.
2. Content velocity is increasing
The localization industry has long dealt with large, text-based projects with lengthy lead times. In fact, even when multimedia localization came onto the scene, the work took place asynchronously, with subtitling or dubbing happening later in the process. Those days are over.
The speed at which companies need to create content has accelerated drastically, and companies face significant pressure to keep up with localization demands.
Everything is immediate. For instance, online customer sales and support require fast-paced localization and interpretation on demand. A growth of online events must accommodate a global audience. Customers expect fast, if not instant, responses no matter what language they speak. So, translation and localization turnaround times have become incredibly short.
As a result of content velocity, the localization industry has become much more complex, a situation that will persist for the foreseeable future.
3. The role of the linguist is changing – not ending
AI is disrupting the industry
It’s long been thought that machine translation (MT) and the larger AI sphere, would disrupt the localization industry, with the most impactful consequences thrust upon the linguist.
While AI and its MT subset will, no doubt, affect the localization industry, forecasts predicting the demise of linguists are simply untrue.
MT is an efficiency tool, much like translation memories, that will help translators perform their work faster. AI will have a greater impact on the entire localization process. You can expect AI to improve localization efforts while translators continue to be an integral part of the process.
Lionbridge is using AI to track the quality of content and to automatically select the perfect translator for the job at hand. Lionbridge clients benefit from a quick and efficient process and can start their translation projects faster.
Expect to see AI and MT play a greater role in the localization industry as time goes on. Changes to the linguist’s job will be something to embrace, not dread.
LSPs have historically performed quality assessments after the content was translated or localized. This process worked well with a traditional asynchronous workflow. It becomes problematic when LSPs are required to pick up the pace, use MT to scale up the workload, and provide deliverables faster.
Lionbridge finds that quality checks must start earlier and apply to the entire process, end to end. The most important dimension of quality is the way the source content is created.
AI can help here too. You can apply AI-based scoring mechanisms to judge a source text to help you determine the next steps for the content. If the content is of low quality, it may not be worth localizing. If it’s of high quality, you may decide it doesn’t need human linguist intervention. AI is used to triage content to determine the best process for each project.
Lionbridge has implemented source quality analysis, which looks at roughly 120 different factors. These include some qualities that have not been considered as much in the past, such as readability and inclusiveness. Then, after conducting MT, human linguists are called in to finesse the text.
The linguists focus on the nuances of meaning and tone. This adds a great deal of value to the outcome. Moreover, these assignments make the translators’ work more interesting. As AI technology continues to take hold, translators can expect more of this type of work.
Outlook for the future
The future will bring exciting changes to the way localization is executed and enable companies to localize more of their content into more languages.
A new content journey is emerging
Organizations must figure out how to deliver a consistent CX across all digital channels and languages. We believe the solution lies in reshaping the content journey. We envisage a content journey that looks like this:
There are activities that localization departments have traditionally handled, and there are other activities that marketing teams have typically handled. Instead of siloed teams carrying out the different requirements throughout the content life cycle, we expect to see companies bring their marketing and localization teams together to create a single global content and customer experience team to support the multimarket strategy of globally-minded organizations.
This reimagined process will enable forward-thinking companies to align technologies and processes and share knowledge and assets. This approach will allow them to enhance the content creation and localization process with smoother operations and deliver better content, faster.
As a result, their localization and content activities will see a higher return on investment.
‘Quality’ will expand and differ by use case
Aspiring to always achieve the “best” quality is no longer the end goal for localization projects of the future. In certain cases, it will be acceptable to produce “good enough” quality.
The increasing demand to deliver more content in more languages at greater speed will create a new definition of quality. This concept of quality will balance linguistic perfection with speed and breadth of delivery.
Lionbridge thinks about this new perception of content quality as a pyramid. On the bottom, we have simplistic content, like user-generated content, which can be of lower quality and localized or produced quite fast. The higher you go up the pyramid, the higher the quality. But this has an impact on the speed and cost of delivery.
There will be multiple quality levels for different types of content, with these various constraints being balanced for each project. Lionbridge will partner with companies to help them better navigate their content journey.
Data will become critical
You need data to effectively leverage AI and machines. Data will be used to feed the algorithms. For instance, to select the right linguist for a job, the machines first need to know something about the content. It needs to be categorized. This means it’s important to provide key information about different aspects like domain, linguistic complexity, and desired quality level. It’s only with this information that the systems will be able to make a match. Data is the key to unlocking this new AI-enabled world.
As companies increasingly rely on localization to motivate customers to act, the localization industry will be at the center of every company’s multichannel experience. And that is something to look forward to following an otherwise challenging year.
Senior Director of Proposition Management, Lionbridge