Working with data that you can trust is a boundary condition for global business environments. The real challenge in processing multilingual or global data is determining with a probability bordering on certainty who you are dealing with and in what form they might appear in any of your data and systems. The answer lies in the application of robust transliteration and transcription, normalization and intelligent comparison methods. This paper offers an insight into the impact of transliteration and transcription on multilingual data matching methodologies. It explains the details with illuminating examples, making the complex issue accessible to a business audience.