Ron Egle is the Content Systems Administrator at Ariel Corporation, headquartered in Mount Vernon, Ohio. He is responsible for identifying workable solutions and maximizing results across the enterprise with touchpoints to enhance the company’s communications, technical documentation, and translation initiatives through digital experiences.
Ron, tell us a bit about what the localization program looked like at Ariel at first.
We started with the simple importing and exporting of Excel files out of our Ektron CMS, and we thought we were doing something really great and automated. But in 2014, we began translating technical content using MadCap Flare and preparing documents for translation using MadCap Lingo. We were working with various language service providers, and were sending Excel files, terminology, and TM files back and forth. A lot of manual back-and-forth going on.
Were there any particular challenges with these manual tasks?
We ended up with our own step of creating in-context reviews, where we built PDFs from the translated content. Of course, the translator vendors offered word reviews and had portals for reviews, but they never saw the content in context. And it got confusing for them. So this whole process was very laborious. And to incorporate those edits into the review, updating the TM, and keeping everything in sync was just a nightmare. So from start to finish, our translation process could take up to 21 steps, and many of them were manual.
Success Stories
How Ariel harnessed extra value from their language assets by using XTM and SYSTRAN
What were some of the real pain points that your company ended up feeling compelled to address in its localization strategy and technology?
The biggest was we had all kinds of translation-memory (TM) quality problems with all the back-and-forth file handling. By manually making review edits, our TM files would be out of order. Furthermore, our regional offices were complaining about our quality problems. We defined terms and had glossaries, but we couldn’t be assured that linguists were actually using them. So that was very frustrating. And our translation costs were high, added to the fact that we weren’t translating materials fast enough to meet our departments’ needs.
What did you decide that you needed?
We decided to replace Ektron CMS with Adobe Experience Manager (AEM). And we determined we needed a translation management system (TMS) to go with it. The TMS had in-context reviews and advanced TM and term base features. It offered us the ability to translate our data projects with all the project-management tools and workflow tools we were looking for. But more importantly, it had a connector that worked seamlessly with AEM. it was able to provide us with a single source, a centralized TMS portal, in which all our translation vendors in our regional teams could work together.
By using a TMS, we’ve reduced the steps needed for translations by 67% and our human translation efforts by 31% . By improving our MT translation quality by 100% using the AI-enhanced TM feature, our language site usage went up from 46% to 84%.”
Ron Egle
Content Systems Administrator
Can you talk a bit more about what features and functionalities you were now using that are key to achieving goals shared by most companies in the manufacturing industry?
It’s not just about having the features, but what’s really nice about a TMS is that they are all integrated into the same tool. So for instance, for our glossary and term base, it’s built into the computer-assisted translation (CAT) tool. As a term comes up, it’s highlighted in the editor for linguists to see and use. That simple, intuitive action that you’d want to have in an editor was there to be used, and that solves the problem of inconsistent terminology, which is key — linguists were now using the proper terms.
And you then looked at further streamlining the localization process. What was the next part is that plan?
We were developing our content-migration plan for AEM, and we determined that we wanted to use machine translation (MT) for lower-level content. We soon realized even greater benefits, as AI-enhanced MT like NFA by SYSTRAN can complete translations without human intervention at a high level of quality by better leveraging our TM. As a result, we’ve reduced our post-editing efforts and increased our productivity, our machine translation quality has improved by 100%.
What have been the overall benefits of implementing all this technology?
Our localization program has reduced the steps needed for translations by 67%. Which is a lot of time savings. We’ve reduced our human translation efforts by 31% and improved our MT translation quality by 100% by using the AI-enhanced TM feature. Our language site usage went up from 46% to 84%. In summary, our content has become way more valuable.