Cutting multilingual formatting time by 63%
Automating the repetitive document work around translation.
The client cannot be identified for confidentiality reasons.
The problem
A body of recurring, template-driven documents produced continually in multiple languages. The translation itself was not the bottleneck. The formatting around it was — preparing source files so they survived translation, and then reassembling structure, styles and layout in every language version afterwards.
Two things made it expensive. It was repetitive, so the same corrections were being applied by hand over and over. And it was unpredictable: the time a job took depended almost entirely on the condition of the incoming file, which nobody could know in advance.
What I built
- Macros that normalize and clean source documents before they go out for translation
- Automated reassembly of structure and styling on the return leg, applied across all language versions in one pass
- Batch operations, so a single correction propagates across an entire document set
- Pre-flight checks that catch the common failure modes before a file reaches a reviewer
The result
Average formatting time fell by 63% — from roughly seventeen minutes a job to six.
The more useful outcome was predictability. Before, an individual job could take up to thirty-five minutes. Afterwards the average was six, and the tail largely disappeared. Capacity planning stopped being guesswork.
The work received an innovation award.
Why it matters
Formatting is treated as an unavoidable cost of multilingual production. A good deal of it is not. The parts that are genuinely craft — judging a typeface for a script, deciding whether to mirror an image in an Arabic layout, catching a rendering fault a reviewer would miss — need a person. The parts that are mechanical do not, and they are usually the majority of the hours.
Knowing which is which takes having done the work by hand for twenty years.
