For Proposal Writers ·
What you'll accomplish
By the end of this guide, you'll be able to upload a full RFP PDF directly to ChatGPT Plus and ask questions about it — no copy-pasting, no text extraction required. Get compliance summaries, evaluation factor analysis, and section drafts based on actual RFP content in minutes.
What you'll need
Log in at chatgpt.com and click "Upgrade to Plus" in the left sidebar. Complete the $20/month subscription. The upgrade takes effect immediately.
What you get: Faster responses, GPT-4o model, file uploads (PDFs, Word docs, images), and DALL-E image generation. For proposal work, file upload is the key feature.
Click "New Chat." In the message input box, look for the paperclip icon (📎) in the left side of the input area. Click it.
Select "Upload from computer" and navigate to your RFP PDF file.
What you should see: The file uploads (usually takes 10–30 seconds for a typical RFP) and appears as a document icon in the message box.
File size limit: ChatGPT can handle PDFs up to ~100 pages well; for larger RFPs, consider uploading individual sections (Section L + Section M as separate files) if the full document is too large.
Troubleshooting: If the PDF is scanned (image-based, not text-searchable), ChatGPT may struggle. Use Adobe Acrobat to "Export to Word" first, which applies OCR, then re-save as PDF.
After uploading, type your question in the same message:
"This is an RFP from [agency]. Please analyze Section M (Evaluation Factors for Award) and tell me: (1) The evaluation factors in priority order, (2) What earns 'Exceptional' or 'Outstanding' for the Technical Approach factor, (3) Three win themes we should build our proposal around."
Press Enter. ChatGPT reads the uploaded PDF and answers based on actual RFP content.
What you should see: A structured analysis that references specific language from the RFP — more accurate than asking without the document.
In the same chat (the PDF context stays active throughout the conversation), ask:
"From Section L of this RFP, list all compliance requirements for the Technical Volume — every 'shall' and 'must' statement. Format as: # | Requirement | Source Paragraph | Risk Level (H/M/L based on how easy it is to overlook)"
Then: "Which requirements are buried in attachments or exhibits rather than the main Section L?"
Ask ChatGPT to draft a section informed by the actual RFP language:
"Based on Section L instructions and the Section M evaluation criteria in this RFP, draft a 500-word Technical Approach introduction that: (1) Opens with a win theme statement, (2) Demonstrates understanding of the government's requirements, (3) Previews our technical approach and differentiators. Use proposal voice: active, benefit-focused, no jargon."
What you should see: A draft that uses language from the actual RFP — not generic GovCon boilerplate — because ChatGPT is working from the document you uploaded.
Continue asking questions in the same conversation — the PDF context remains: