Reference documents provide the AI with concrete examples of content, structure, and language style. Instead of merely describing to the AI what the result should look like, they use previously approved documents to show which wording, structure, and level of detail are required.
This helps the AI understand the desired context better and create new content based on existing company documents.
Suitable reference documents include, for example:
- Already published content: Blog posts, newsletters, or customer communications.
- Technical documentation: User manuals, data sheets, or product specifications.
- Marketing and sales materials: Brochures, presentations, or case studies.
- Company documents: Policies, process descriptions, or training materials.
An example:
❌ Without a reference document:
“Create a product description for our new module.”
The AI independently chooses the structure, writing style, and level of detail.
✅ With a reference document:
“Create a product description for our new module. Use the attached product description of our existing module as a guide for the structure, level of detail, and language style.”
The new product description follows the style of the existing documentation and integrates seamlessly into the existing product range.
In short:
Reference documents provide the AI with a concrete model and help create content that matches existing company documents in terms of subject matter, language, and structure.