Global market leader of electronic components distribution slashed time to market with Unbabel by adopting Post Edited Machine Translation (PEMT).
Highlights
Unbabel built a bespoke Machine Translation (MT) engine, which reduced translation time without affecting the translation quality. Using MT helped the client stretch its translation budget further, unlocking more content for multilingual markets. The PEMT approach allowed the client to reduce their time to market internationally from four weeks to one, enabling them to grow their global market share by 40%.
About Electronic Components Distributor
This electronic components distributor is a global market leader:
serving over 1100 industry-leading suppliers;
conducting business across 27 currencies;
providing localized support from 25 locations around the globe.
The Challenge
The client had a good reputation for launching products faster than the competition in their home US market. Speed to market was key to capturing its competitive advantage, but it was taking four weeks to release products in global markets.
It was losing its first mover advantage. The client needed to accelerate its localized product launches across its international markets.
The Solution
Consultation and assessment
Unbabel spent time with the client to unpick its global business objectives and carried out a localization strategy review. This assessed the end-to-end content process for international launches, from content creation to translation and review. The Unbabel team identified opportunities for improvement and devised a plan for achieving the client’s global business objectives through a strategic localization approach.
Three key actions were identified to reduce time to market:
create customized MT engines from the large batch of previous translations;
commence PEMT to make the translation phase faster;
establish client review of the content to reduce review time.
Post-Edited Machine Translation (PEMT) is the combination of high-quality machine translation (MT) and human post-editors. Post-editors only need to review and edit previously-machine translated content. For the best outcome, Unbabel fine tunes PEMT engines to our clients’ content, and our state-of-the-art adaptive engines learn from the reviewer’s edits in real time.
PEMT is key to localizing at scale; it allows new content to be quickly translated and can be a cost-effective alternative to the human-only translation process, helping clients do more with their translation budget.
Commence Post-Edited Machine Translation
Translation quality was key to the client. So, to use PEMT as an accelerant, the final output had to meet their exacting requirements.
Unbabel began with a trial approach testing a single language, building an engine with the client’s existing language assets so it was tuned to its exact content. The feedback and productivity metrics from the translation platform showed that the custom-built engine was providing the same quality, in less time than traditional human-only translation.
The Outcome
Enhanced business agility: Time to international markets was reduced from 4 weeks to 5 days.
Increased revenue: The client saw a 40% global market share growth.
Cost reduction: Abandoning translations from scratch in favor of customized PEMT drastically cut costs.
Faster translation process: Using a customized MT engine and introducing an in-country review saved time and resources.
More content with the same budget: By reducing the translation costs, the client could extend its international product lines.
Boosted confidence in the quality: Pairing customized engines with experienced editors resulted in significantly better content than from generic engines available on the market.
Constant real-time translation improvement: Rather than needing to be retrained, the engine learned in real-time from edits and improvements made by reviewers, resulting in more on-brand content.
Are your localization efforts up to par? Find out now with our localization scorecard.
