How to Tell AI Exactly How to Format Its Answers (RTF Formula)
- JB
- Mar 3
- 3 min read
Updated: May 18
If you have been following along with the RTF Formula series here on try-AI, you already know that a well-built prompt has three parts: Role, Task, and Format. We have covered Role (telling AI who to be) and Task (telling it what to do) in a different article. Now it is time to talk about the piece that most beginners skip entirely: Format.
Format is the instruction, usually at the end of your prompt, that tells AI how to package and present its answer. You're not changing what it says. You are just telling it how to hand the information to you. And that one small shift makes the difference between an answer that's immediately helpful and one you you have to go back and clarify. Or give up on.
Why Format Matters More Than You Think
Here is what happens when you skip the Format step: AI guesses. And its default guess is almost always to write in paragraphs. Sometimes lots of them, and completely overwhelming if all you wanted was a quick dinner idea or a simple checklist. Other times, AI will ask you a bunch of other questions to arrive at the output you should have been clear about from the start.
According to research from Stanford's Human-Computer Interaction Group, people process structured information like lists and tables much faster than unstructured writing. So when AI spits out information that is laid out clearly, your comprehension improves and you're better able to digest what you're reading. This matters whether you are reading a blog post or an AI response.
So, the fix isn't complicated. You just need to know a handful of Format instructions and when to use each one. Here are the six most useful ones for everyday life.
The 6 Most Useful AI Format Instructions
How to Actually Use These in a Real Prompt
The Format instruction always goes at the end of your prompt, after you have told AI what you need. Here is a simple example using all three parts of the RTF formula together:
The difference in output is significant. One comes back as a wall of text. The other comes back as something you can actually tape to your fridge.
A Note on Mixing Format Instructions
You can absolutely combine more than one Format instruction in a single prompt. For example: "Give me a numbered list, keep it under five items, and use plain language." AI handles layered instructions well as long as they do not contradict each other. The more specific you are about how you want the answer, the more useful the answer will be.
This is something I had to learn through experimentation. When I first started using AI tools, I treated them like a search engine. I would type a question, get back a wall of text, and assume that was just how it worked. It was not until I started being deliberate about Format instructions that things started clicking. A prompt with a clear Format instruction stopped feeling like a guessing game and started feeling like a conversation with someone who actually listened.
That shift is exactly what try-AI is here to help you experience for yourself.
The Bottom Line
You do not need to master prompt engineering to get dramatically better results from AI. You just need to be more specific about how you want the answer delivered. The six Format instructions above cover the vast majority of everyday use cases. Pick one, try it on your next prompt, and notice the difference.
Role tells AI who to be. Task tells it what to do. Format tells it how to hand you the answer. All three together? That is when AI starts feeling less like a research tool and more like a personal assistant who actually gets it.
Watch the Full Video
This article is a companion to the try-AI video: "Stop Asking AI for Walls of Text. Do THIS Instead." Watch the full video for live screen demos of Format instructions in action, including a side-by-side comparison of prompts with and without the Format step.
References and Further Reading
Stanford HAI. (2023). Human-Centered AI. Research hub on the intersection of AI outputs and human comprehension.
OpenAI. (2024). Prompt Engineering Guide. Official documentation on structuring prompts for better AI outputs.
Anthropic. (2024). Claude Prompt Engineering Overview. Guidance on formatting and structuring prompts for Claude.
Mollick, E. (2024). Co-Intelligence: Living and Working with AI. Wharton professor Ethan Mollick's accessible guide to working effectively with AI tools in everyday and professional contexts.




Comments