Empower Your Marketing Team: Promoting the Use of Machine Translation with the MT Suitability Score

Track: Multilingual AI | AI1 |   Everyone |
Wednesday, June 4, 2025, 9:00am – 9:30am
Held in: High Live 1 (Plenary)
Presenter:
Vera Malysheva - TeamViewer
Host: Brona Nilsson

The biggest challenge when implementing machine translation on an enterprise level is expectation management. Colleagues without a linguistic background can be both too eager and too hesitant, resulting in uncertainty around appropriate use cases. Rather than create complex guidelines, our Localization team came up with a customized risk-based scoring model based on TAUS DQF that allows requesters to score their text against 4 categories: Utility, Sentiment, Time and Longevity. The result is a number between 0 (not suitable) and 10 (very suitable), empowering the requester to decide the level of quality required and raising risk awareness.

Key Takeaways:

Quality is a subjective metric; risk is an important factor when considering machine translation; localization teams have the power to educate and prove their value by helping adopt language AI