Prompt Engineering Basics: The 4-Part Formula

By the AI Playbook team · Beginner · 25 min read · Last updated: October 13, 2026

The difference between a vague AI answer and a perfect one is almost always the prompt. This tutorial teaches one reusable formula — Role, Task, Context, Format — that works across every AI tool.

RoleWho the AI is TaskWhat to do ContextBackground info FormatHow to present it The 4-Part Prompt Formula
Diagram: Role, Task, Context, Format — the four ingredients of a reliable prompt.
  1. Learn the 4 parts of a good prompt

    Every strong prompt answers four questions:

    PartQuestion it answersExample
    RoleWho should the AI act as?"Act as an experienced nutrition coach…"
    TaskWhat exactly should it do?"…create a 7-day vegetarian meal plan…"
    ContextWhat background does it need?"…for a beginner who hates cooking and has a $40 weekly budget…"
    FormatHow should the answer look?"…as a table with columns for day, meals, and shopping list."
  2. Rewrite vague prompts using the formula

    Watch the difference. Weak: "Write something about productivity." Strong:

    Try this prompt
    Act as a productivity coach for freelancers. Give me 5 specific morning habits that improve focus. I'm a writer who gets distracted by social media. Format: a numbered list, each habit one sentence plus one sentence explaining why it works. Keep the total under 200 words.

    Practice: take any prompt you've written before and add the missing parts. Most will be missing context and format.

  3. Use constraints and examples

    Two power moves: constraints (word count, tone, "no jargon") and giving an example of what you want. Example: paste one of your old emails and say "Match this tone." AI copies style remarkably well when it can see it.

    Try this prompt
    Write 3 subject lines for a newsletter about home gardening. Constraints: each under 8 words, curious but not clickbait, no exclamation marks.
  4. Iterate with follow-ups

    Treat the first answer as a draft. Useful follow-ups: "Make option 2 longer and drop option 3." · "Rewrite for a 10-year-old." · "Give me 3 more variations in a funnier tone." One round of iteration typically improves output quality more than rewriting the original prompt.

  5. Build a personal prompt library

    Save prompts that worked well — a notes app or document is fine. Tag them by task (emails, planning, studying). Within a month you'll have a personal toolkit that turns 10-minute tasks into 2-minute tasks. Revisit and improve entries as you learn.

10 ready-to-use prompts to practice with

  1. Summarize this article in 5 bullet points for a busy person.
  2. Act as a travel planner: 3-day budget itinerary for [city], under $300 total.
  3. Rewrite this paragraph to sound warmer but keep every fact. [paste text]
  4. Explain [topic] using one everyday analogy, then test me with 2 questions.
  5. Act as a job interviewer: ask me one tough question at a time, then give feedback.
  6. Brainstorm 10 blog post titles about [topic]; make 5 practical, 5 controversial.
  7. Proofread this text, fix grammar, and explain the 3 biggest errors. [paste text]
  8. Act as a math tutor: solve step-by-step, then give me a similar problem to try.
  9. Compare these two options as a pros/cons table. [paste options]
  10. Turn my rough notes into a polished 150-word LinkedIn post. [paste notes]

What it costs

Prompting skill costs nothing — it works on every free AI plan. Paid plans (around $20/month) give you stronger models that follow complex instructions better, but the formula matters far more than the model. A great prompt on a free model beats a weak prompt on a paid one.

Can this make money?

Realistic take: prompt skill itself is now common knowledge — it's a multiplier, not a product. It makes you faster at writing, marketing, coding, or teaching. Anyone selling "secret prompt packs" as a shortcut to riches is selling the shovel, not the gold. Learn it, apply it to a real skill, and skip the hype.