File Formats Used in Translation: A Practical Guide for Businesses

A 5,000-word Word document and a 5,000-word InDesign file may contain exactly the same number of words, but the workflows, tools, quality checks and delivery requirements can be completely different. Professional translation, at present, involves far more than exchanging Word documents. Translators and localization teams work with websites, software, spreadsheets, presentations, design files, subtitles, structured data and content management systems (CMS).

Why File Format Matters in Translation

A translation team needs to work with the actual content while preserving as much of the original structure and formatting as possible.

The file format determines:

• How text can be extracted
• Whether formatting can be preserved
• Whether CAT tools can process the content
• How placeholders and variables are handled
• Whether images contain text
• How the translated content will be delivered
• Whether desktop publishing or technical work is required
• How easily the translated content can be integrated back into the product

For example, translating an editable .DOCX file is relatively straightforward. A scanned PDF containing images of printed pages is an entirely different proposition. And a JSON file containing thousands of software strings is yet another beast humanity has decided to create.

1. DOCX: Microsoft Word Documents

DOCX is one of the most common formats in professional translation.

It is widely used for Contracts, Reports, Manuals, Certificates, Business correspondence, Marketing & Legal documentation etc.

DOCX files are generally well suited to CAT tool workflows, as text can be extracted while much of the document's structure is retained.

Potential issues: Complex Word files may contain Text boxes, Embedded objects, Tables, Headers and footers, Hidden text etc.

2. PDF: The Popular Format

PDF is probably one of the most frequently exchanged document formats in the translation industry. It is primarily designed to preserve document appearance, not necessarily to make the underlying text easy to edit.

Text-based PDF

The text can be selected and extracted. This can often be processed efficiently, although complex layouts may still require additional work.

Scanned PDF

The document consists primarily of images. In such cases, the translation workflow may require OCR → Text Extraction → Translation → Formatting/DTP → QA.

OCR quality can also affect translation quality. A badly recognised character can become a badly translated word. For certified documents, this distinction becomes particularly important because names, dates, stamps, signatures and official terminology need to be handled carefully.

3. XLSX: Excel Spreadsheets

An .XLSX file includes Cell content, Formulas, Hyperlinks, Dropdown menus, Metadata etc.

Therefore, professional translation workflows need to distinguish between translatable content and content that must remain untouched.

4. PPTX: Presentations

PowerPoint presentations are another common translation format, used widely for Corporate presentations, Conferences, Pitch decks, Internal communication about products/services etc.

A major challenge with PowerPoint localization is text expansion. Translated text may occupy more or less space than the source.

This can result in text overflowing a text box, changes in line breaks, distorted layouts etc. Consequently, PowerPoint translation often requires a visual QA and formatting review after linguistic work.

5. IDML and InDesign Files

Marketing departments, publishers and design agencies frequently use Adobe InDesign for Brochures, Magazines, Flyers and Catalogues.

The native InDesign format is .INDD, but IDML is often used in translation workflows because it provides a more exchange-friendly format for CAT-tool processing.

Typical workflow: InDesign → IDML → CAT Tool → Translation → Linguistic QA → DTP → Final PDF

The important distinction here is that translation and graphic design are not necessarily the same task. A linguistically correct translation can still require desktop publishing work to make the final document visually correct.

6. HTML and Website Content

Website localization frequently involves HTML files or content exported from a CMS.

For example:

<h1>Welcome to our website</h1>

The visible text may need translation, while the HTML tags should remain intact.

Professional workflows therefore need to protect:

• HTML tags
• Links
• Attributes
• Variables
• Metadata
• Formatting elements

7. XML

XML is commonly used for structured content and data exchange.

It appears in Technical documentation, Product info, Software, CMS, Enterprise applications etc.

XML separates content from structure. This means the translator may work with the textual elements while technical tags remain protected.

A simplified example might look like:

<title>Installation Guide</title>

The text may be translated, while <title> remains unchanged.

Because XML structures can vary significantly between systems, translation teams need to understand the specific XML schema and rules involved.

8. JSON

JSON is particularly important in modern software localization.

It is commonly used in case of Web & Mobile Apps, SaaS platforms, APIs and Software interfaces.

In this format, professional localization teams must pay special attention to Keys, Placeholders, Variables, Syntax, Special characters etc. One misplaced character can turn a translation problem into a software problem.

9. XLIFF

XLIFF, or XML Localization Interchange File Format, was designed specifically for exchanging localization data.

It is commonly used between CAT tools, TMS, Software development environments and Localization platforms.

XLIFF can contain source and target segments, metadata and other information required during localization. Because it is designed around localization workflows, it is particularly useful when content moves between different systems.

10. CSV and TXT Files

Simple formats such as .CSV and .TXT are also used in professional translation.

CSV files can contain Product descriptions, Keywords, Metadata, Database exports etc.

However, CSV files require careful handling because columns may have different functions.

Translation professionals need to know which fields should be translated and which ones should remain untouched.

11. PO Files

PO (Portable Object) files are frequently encountered in software and website localization, particularly in open-source and multilingual software environments.

They contain source strings and their corresponding translations. A typical PO workflow involves translating the relevant message strings while preserving the technical structure of the file. This is another example of why professional localization requires knowledge beyond traditional language skills.

12. SRT and VTT: Subtitle Files

Multimedia localization has its own formats.

Two common subtitle formats are .SRT and .VTT.

They contain Subtitle text, Timing info, Sequence info.

Apart from linguistic accuracy, subtitle translators need to consider reading speed, synchronization, character limits, line breaks and speaker identification, among other things.

13. Images and Scanned Documents

Images like posters, ads, infographics, flowcharts, screenshots, product packaging etc. can also contain translatable content.

These often require Image → OCR → Translation → Graphic adaptation → QA.

The final deliverable may need to reproduce the original design while replacing the source-language text with the target language. This is where translation, localization and design intersect.

Editable vs. Non-Editable Files

One of the most important distinctions for clients is whether the source file is editable.

Usually easier to process

• DOCX
• XLSX
• PPTX
• IDML
• HTML
• XML
• XLIFF
• CSV
• SRT/VTT

Potentially more complex

• Scanned PDFs
• Images
• Screenshots
• Flattened artwork
• Low-resolution documents
• PDFs with complex graphical layouts

This doesn't mean that difficult formats cannot be translated.

It simply means the workflow may involve additional steps such as OCR, extraction, DTP, engineering, formatting or visual QA.

File format is not just a technical detail. It is part of the translation workflow.

A professional translation project needs to consider the following:

What is being translated?

Where will it be used?

Which parts of the file are translatable?

What needs to remain technically untouched?

How will formatting be preserved?

How will the final content be tested?

The answers determine the appropriate workflow, tools, resources and QA process.

From a simple DOCX document to a multilingual SaaS platform, professional translation is ultimately about more than converting words from one language into another.

It is about delivering usable, accurate and contextually appropriate content in the format your audience actually needs.