Everyday Business Prompt Frameworks: CREO, PECRA, GRADE and More

Study guide

Smaller Prompt Frameworks and When to Use Them

This course covers eight prompt frameworks that sit outside the well known RTF and CO-STAR: CREO, PECRA, GRADE, PARA, ERA, RISE, ACE and CTF. Each one is a different mix of the same six ingredients, which are role, task, goal, context, format and example. Once you can see those ingredients, the acronyms stop being intimidating. You will learn what each framework adds, what it leaves out, and which jobs it suits. The course is written for busy professionals who paste prompts into ChatGPT, Claude or Gemini and want answers that fit the real work, not generic text they have to rewrite.

The teaching is practical throughout. Every framework is explained with the why, the how, a worked example from a normal job, and the mistake that most often spoils it. You will see why CREO and PECRA put the reason first, why GRADE ends with a sample to copy, why PARA only works after something has been solved, and why RISE, ERA and CTF are best kept for small jobs with a known procedure. By the end you will not be collecting frameworks. You will have one quick framework for short asks, one complete framework for bigger tasks, and a clear rule for borrowing the most useful slots when a prompt needs more clarity.

What is a prompt framework?

A prompt is the request you type into an AI tool such as ChatGPT, Claude or Gemini. A prompt framework is a reusable structure for that request: a short checklist of the parts a good prompt contains, usually named with an acronym so it is easy to remember. Why it helps: the AI fills in whatever you leave out with generic guesses. A framework makes sure you include the parts that change the answer, such as who it is for, what you want to achieve and the format you need.

Frameworks make prompts repeatable: you fill the same parts each time, save the prompt as a template and share it with colleagues. They are not magic words: the value comes from the detail you put in each part. They help most when a request is complex or open to interpretation; a quick, simple question can stay short. There are well over sixty named frameworks, and most of them remix the same ingredients: role, task, context, audience, format, constraints and examples.

In this course you learn Business frameworks, which is a fill-in-the-slots framework: an acronym where each letter stands for one part of your prompt. It is best for: professionals who want to know the lesser-known frameworks and which ones add something.

Many names, same ingredients

Dozens of prompt frameworks are in circulation, and most of them are rearrangements of six core ingredients: role, task, goal, context, format and example. RTF covers role, task and format. CO-STAR adds context, audience and tone. Treat those two as your baseline. When you meet a new framework, the useful question is simple. What does it add that RTF or CO-STAR does not? If the answer is nothing, you can safely skip it and keep using what you already know.

This habit also helps when you read a prompt written by a colleague. Instead of hunting for the acronym, look for the ingredients. You will quickly see whether the role is missing, whether the goal is buried, or whether there is no format instruction at all. That diagnosis is usually enough to improve the prompt without rewriting it from scratch.

The common mistake here is collecting frameworks like trading cards. Learning a new one every week feels productive, but it rarely improves your output. Two or three frameworks that fit your actual work will carry you much further than a long list you never use properly.

CREO and PECRA: purpose first

CREO stands for Context, Request, Explanation, Outcome. The Explanation slot is where you say why you need the answer, and it is the part most people forget. PECRA stands for Purpose, Expectation, Context, Request, Action. It puts Purpose first, before any background, so the reason leads the whole prompt. Both frameworks come from Fabio Vivas, and both are built for requests where the why changes the answer, such as research briefs and decision support.

The choice between them is about order. Use CREO when you already know the situation and want to add the reason to an otherwise normal request. Use PECRA when the reason itself is the most important thing to state first, because everything else should be read in its light. A quick factual lookup does not need either framework. Save them for work where a generic answer would waste your time, like a brief that informs a meeting or a decision.

A worked example makes this concrete. Suppose you need a summary of three supplier quotes before a budget meeting. In CREO you would set the context of the quotes, make the request for a summary, explain that the meeting decides next year's contract, and state the outcome you want, which is a one page comparison with a recommendation. The explanation is what stops the AI from producing a neutral list with no view on which quote is best value.

GRADE and PARA: goal and application

GRADE stands for Goal, Request, Action, Details, Example. It is a goal led cousin of CARE, which means the goal is stated before anything else so the AI filters every later line through it. The Example slot is what separates GRADE from most other frameworks. When you have a past email, report or slide that worked, paste it in and the AI copies its shape instead of guessing at a style you never described.

PARA stands for Problem, Approach, Result, Application. It suits technical write ups and case studies because it follows the natural story of something that was already solved. Use GRADE when you have a sample to share. Use PARA when you need to tell the story of a solved problem. The deciding question is simple. Can you point at an existing document, or do you need to explain what happened?

Two failures are worth remembering. The first is burying the goal at the end of the prompt. The AI reads top to bottom, so a goal placed last competes with everything above it. The second is using PARA for a brand new task. If there is no solved problem yet, the Problem and Result slots have nothing real to hold, and the AI fills them with invented content that looks plausible and is wrong.

RISE, ERA and CTF: the short ones

RISE stands for Role, Input, Steps, Expectation. It is the predecessor of RISEN, which adds a Notes slot at the end. Use it when the job follows a known procedure, because the Steps slot lets you fix the order of operations. ERA stands for Expectation, Role, Action, and it puts the expected result first. That order helps when the shape of the output is the difficult part, because the AI reads your expectation before the instruction. CTF stands for Context, Task, Format. It is RTF with context instead of role. RTF asks the AI to play a person, while CTF tells it what situation it is working inside.

Short frameworks are a fit, not a virtue. If a job has several audiences, formats or stages, three slots will not hold it, and you should move up to PECRA or GRADE. The Steps slot in RISE is easy to skip, but skipping it lets the AI invent its own order. For data work, the order is often the part that matters most, because a wrong sequence produces a tidy answer built on the wrong calculation.

In CTF, the Context slot does the heavy lifting. A clear background about where the data came from and what is wrong with it replaces the role you would have written in RTF. For example, telling the AI that a sales export covers two regions, uses different currency formats and includes test orders is more useful than asking it to act as a senior analyst.

ACE: audience first

ACE stands for Audience, Context, Execution. It is used for brand consistent marketing, where the reader and the tone matter as much as the facts. Starting from the audience keeps text focused on what the reader needs, instead of describing what the company did. The same product update reads very differently when written this way, because the reader's situation decides which details are worth including.

ACE overlaps with CO-STAR. Both ask for context and both shape the output. Pick ACE when you want something shorter and the reader is the whole point. The Context slot carries the facts the reader already has and the situation they are in. Leave it out and the model fills the gap with guesses. The Execution slot sets format, tone, length and the next step. Without it you often get a wall of text with no clear action.

The audience line is where most ACE prompts fail. A weak one uses words like everyone, users or people. A strong one names a single reader with a real situation, such as a small business owner on the starter plan who has just hit the limit on monthly invoices. That level of detail changes the vocabulary, the length and the call to action without any further instruction.

Picking from the long tail

You do not need all eight frameworks. Pick one quick framework for short asks and one complete framework for bigger tasks, then stop there. Borrowing slots is allowed and often sensible. CREO's Explanation, PECRA's Purpose and GRADE's Example can be added to any other framework when a prompt needs more clarity. The acronyms are guides, not rules.

A shared framework makes prompts reusable. When everyone uses the same shape, a prompt written by one person can be copied, improved and passed on. Standardising on one template turns prompting from a personal trick into a team habit, and an improvement to a single slot benefits everyone who uses the template. That is where the real return comes from, not from learning another acronym.

The biggest risk is framework collecting, learning a new one every week and never settling. Two frameworks used well beat eight frameworks used once. The whole course reduces to one idea. Eight frameworks share the same ingredients. The names differ, the slots differ, the job is the same.

Frequently asked questions