PDFs That Just Won’t Go Away – Or Maybe They Can?

Let’s talk about those pesky PDF files that always show up needing extra TLC before and after translation. Are we beating a dead horse here by bringing the topic of PDFs up again? 

Absolutely not. The horse is still alive and galloping.

In fact, PDFs are like old Volvos: they never really die.

_{In this post, we’ll examine how our team tackled the challenge of translating scanned PDF forms with handwritten fields—without mentioning AI (promise!) and by sticking to a good old-fashioned, practical workflow.} 

The Story in Short 

1. Background 

• Large Pharma Client: We work with a major pharmaceutical company that often provides batches of scanned PDF files for translation. 

• Document Specs: Production sheets that are complex forms to be filled in on the factory floor. Some PDFs can be up to 80 pages, with about 350 words per page. 

• Template-Based: The original documents were created in MS Word with a lot of boilerplate text. 

• Mixed Text Inputs: Certain fields were pre-filled in Word, while others were handwritten on printed copies. 

• Scanned PDFs: These annotated hardcopies were finally scanned to PDF and sent over for translation. 

2. First Shot – Preliminary Workflow and Estimate 

• OCR Extraction: Our initial approach was to convert the PDFs to Word via OCR to facilitate translation. 

• Handwriting Cleanup: We would then verify and manually type in the handwritten content if OCR missed or misread it. 

• Boilerplate Check: Even the boilerplate text needed to be vetted for OCR errors. 

• Time Estimate: We calculated about 240 hours of pre- and post-processing to make the content TMS-ready and then finalize the layout after translation. 

• Outsourcing: Given the sheer volume, outsourcing was an option. However, our project manager realized we’d still need to thoroughly review the vendor’s work due to the complexity. That essentially defeated the purpose of outsourcing, so we considered a different route. 

3. Calling In Our Black-Belt DTP Team 

• Leverage Original Word Files: Since most text was originally generated from Word templates, our experts opted to translate the native Word documents directly in the TMS, thus benefiting from translation memory and avoiding needless rework. 

• Isolate Handwritten Fields: Any handwritten fields were scanned, OCR-processed, and checked separately. Translations were then manually inserted back into the final Word files. 

• Time & Cost Reduction: This approach achieved 25% cost savings compared to the original plan—no more extensive boilerplate fixing and layout headaches. 

• Added Consistency: Our team also introduced field references in Word for key data to further streamline updates and ensure consistency. 

4. A Key Enabler of Success 

• Client Collaboration: The pharma client’s willingness to provide all the original Word documents was crucial. Their understanding of PDF limitations allowed us to create a cleaner workflow. 

• Final Format Delivery: Although the client still requested PDF deliveries, they recognized the advantages of working in native MS Word. 

_{Take it straight from the horse’s mouth: that’s how you process content in a lean and efficient way!} 

‍