ChatGPT Projects for Beginners: 5 NEW Use Cases

About this video

David (presenting for the channel) shows how to use ChatGPT Projects to handle repetitive tasks consistently. He notes Projects appear in the left sidebar but, at the time of recording, were not yet available on the free plan despite OpenAI’s promise, so a paid plan is needed to follow along. After creating a ‘riddles’ project, he shows organizing chats by dragging them in or via a chat’s three-dot menu, and adding project-wide custom instructions — a demo instruction set forces bullet-only answers at a sixth-grade reading level starting with a fun emoji, and it applies to every chat in that project only. Five use cases follow: (1) a mentorship project where uploaded documents of product-building principles are added as files and instructions require advice to be tied back to the originating principle; (2) weekly chart-building, where instructions specify total revenue plus a horizontal bar chart by store and a vertical bar chart by product line from an uploaded PDF sales report, refined by trial and error (he warns bar widths varied, so put consistency requirements in the instructions); (3) data entry converting insurance-plan PDFs to a specified CSV format, downloadable from the result; (4) customer-feedback analysis from a CSV, using a markdown template with bracketed placeholders to produce an action plan identifying the most common negative themes; and (5) negotiation emails for discount requests, with instructions to write in a negotiate-but-don’t-agree, solution-oriented style. He closes by comparing Projects with custom GPTs: Projects are simpler to create, allow more models (including a reasoning model), but cannot be shared and chats inside a project do not share context with each other; custom GPTs can use Actions to connect external services and can be shared by link.

Key takeaways

• Project custom instructions and uploaded files apply to every chat in that project, so set format, level and rules once instead of re-prompting
• Turn recurring reports into an upload-only workflow: specify the exact outputs (totals and chart types) in instructions, then just drop in each new PDF
• Specify the exact CSV columns/format in instructions to convert PDF data-entry work into a downloadable table
• Give feedback analysis a fixed markdown output template with placeholders so each run returns a comparable action plan focused on the most common negative themes
• Projects cannot be shared and their chats do not share context with each other; use a custom GPT instead when you need sharing or external-service Actions

Episode details: published March 6, 2025 · 10:42 long. Summary written by Self Growth Videos from the episode itself. Business results and any income figures mentioned are Kevin Stratvert’s own claims, shared for education — they are not typical or guaranteed.

← Previous: 7 Best AI Tools You NEED to Try (Free & Powerful!) 💡 · Next: How to Create Ads with AI In 15 Minutes (INCLUDES Variations!) →

All Kevin Stratvert videos · Kevin Stratvert profile · Kevin Stratvert on YouTube