A short presentation describing a prompt/workflow strategy for improving ChatGPT outputs by using multiple staged ChatGPT contexts: separate research-context, prompt-design, and execution windows. The speaker claims this multi-window approach (plus a campaign agent with brand and campaign context) yields substantially better creative outputs and automates tasks like email templates and design mockups.
Chapters
If you were just opening up chatgpt and typing stuff in you're going to get an okay result. If you are doing a deep research project using that deep research project as context for ChatGPT you're then opening a chat GPT window and working with it on a prompt and then you take that prompt and that deep research to generate an outcome. So you're doing three chatgpt windows instead of one. You're going to be 10 to 20 times more successful in terms of quality of output. Right? Exactly. And now I have a campaign agent that I can then have make any asset in this campaign I want right? I can say I want it to mock up an email template. Oh can you mock an email template design for this campaign? Pretty incredible because I have all of the brand knowledge I have all of the campaign context and now I can just like I can just basically jam. It's my creative director and art department for me right? Is that not wild that you mind blowing.