Welcome to the C.R.E.A.T.E Formula for Prompt Engineering, a succinct supplement to the full Prompt Engineering A-Z guide you received in our Welcome email. This mnemonic-based tool is designed to streamline your understanding of prompt engineering, connecting the dots from the detailed content you've already received. Let's dive into C.R.E.A.T.E and simplify this process further.
An Easy Mnemonic: C.R.E.A.T.E
C: Character - Tell ChatGPT what role you want it to play. You are letting it know its role and motivation. Ex: “You are a talented copywriter with 20 years experience writing high-performing sales copy” ⇒ Not Zero, Drawing the best on Sales Copy
R: Request - Tell the AI what we want it to do. I recommend that you’re specific here. It’s not just “Write me a sales email for a sports car” ⇒ “Write me a persuasive and enthusiastic sales email for the Cougar Hyper Sport. This is a high-performing electric sports car with the best acceleration in its category. The sales letter should tell people they need to experience it for themselves, find out what it feels like to be pinned to your seat, to feel the heart-thumping exhilaration of cornering, and to enjoy the experience of everyone looking to see who’s in the driving seat. ” This is what defines the quality of the response.
E: Examples - Optional - If you want to give ChatGPT some better direction, it might be worth adding some examples. Maybe some examples of previous headlines, or an example of the tone of the voice. ChatGPT is great at learning from these and delivering a response in a similar style.
A: Adjustments - When you tried a prompt, and it’s not quite giving you exactly what you want as a response. These are the refinements. For me, they’re often phrases like: Don’t use bullet points, use subheads to break up the text.
T: Type of Output - Tell ChatGPT exactly how you want it to format its response. Ex: Deliver your response as a 500-word article with a headline and a conclusion. The type of output can include articles, bullet points, tables, poems, jokes, scripts, anything that can be written.
E: Extras - Little additions that can really make a big difference. Ex 1: Ignore everything before this prompt/ Ignore all previous instructions before this one - This is useful if you don’t want to draw on any of the conversations that you’ve had up until this point. I often use this at the beginning of a prompt when I’m changing tasks Ex 2: Ask me questions before you answer. I often use a variant on this when I know that the AI needs lots of information to get its response right. We’ll then have a conversation where I answer lots of questions and when ChatGPT thinks it got enough information. It will then deliver a response ⇒ This one was an absolute game-changer for me.
Ex 3: Explain your thinking/What do you think?. It’s the equivalent of showing you’re working in maths class. It’s a way of verifying an answer by making sure the journey to that answer makes sense. This one is super useful.
Ex 4: Step by step:
Ex 5: Summarize in a paragraph or two everything you know about me. Write in the first person.
Ex 6: Your answer must be use your database/insights/information upto 2023.
Ex 7: Answer only using reliable sources and cite those sources.
Ex 8: Use Forceful language to emphasize something, or your request - CAPLOCK Everything.
If you follow the C.R.E.A.T.E Formula, you’re going to start writing prompts they are far more effective. I admit that they will take a bit of time to write, but the extra effort pays off hugely in the results.
****I hope you enjoyed this guide, don’t forget, if you have any questions, reach out to us at [email protected] & ****join our newsletter to get all our upcoming guides and tutorials and see more real-world examples and case studies so you can practice more and more!