4 universal formula - AI Prompt Engineering

Learning Objectives

Learn the 7-step universal formula for prompt construction.

Understand the correct order to present information to an LLM.

See how applying the formula transforms a simple request into a professional prompt.

The 7-Step Formula

Now that you know the anatomy, here is the universal formula that works across ChatGPT, Claude, Gemini, and almost any other LLM.

Whenever you are about to write a complex prompt, construct it in this exact order to ensure the AI parses your instructions with maximum logical weight:

Set the Role

"Act as an expert copywriter..."

Define the Task

"Write a landing page for my new SaaS product."

Provide Context

"The product is a habit tracker designed specifically for students with ADHD."

Specify Audience

"The target audience is college students aged 18-24."

State Constraints

"Do not use complex corporate jargon. Keep sentences under 15 words."

Set Tone

"Use an empathetic, encouraging, and highly motivating tone."

Dictate Format

"Format the output as a Markdown document with clear H1 and H2 headers."

Why Order Matters

Language models process text sequentially (autoregressively). By putting the Role first, you immediately set the statistical weights of the model's vocabulary. If you put the role at the very end of a long prompt, the model has already processed the context using a generic persona.

The Prompt

Write a landing page for my habit tracker app.

Analysis

The AI will generate a generic landing page assuming standard features and audiences.

Knowledge Check

Why is it recommended to put the 'Role' at the very beginning of the prompt rather than at the end?

Executive Summary

The Universal Formula (Role, Task, Context, Audience, Constraints, Tone, Format) provides a predictable structure that works across all major LLMs.

Up Next

Lesson 02

Writing Your First High-Quality Prompt

15m Beginner +50 XP

Continue

Previous: The Anatomy of a Perfect Prompt