The big idea
A good prompt is a brief, not a wish — it has up to five parts: role, context, task, format, and examples. Context is the single highest-leverage ingredient because the model can infer tone and structure but can't invent facts about your business, and the second prompt (your iteration) usually matters more than the first.
Deep dive
The mental model — brief it like a freelancer
Imagine you hired a brilliant freelancer who has read the entire internet but knows nothing about you, your company, your customers, or your taste. You wouldn't email them "write me a blog post" — you'd send a brief. Every good prompt is a brief with up to five components, and when output disappoints it's almost always because one of these is missing: **Role** (who should the AI be), **Context** (what it needs to know — your company, audience, goal, constraints), **Task** (exactly what to do, with a specific verb and deliverable), **Format** (length, structure, tone, a template to follow), and **Examples** (what "good" looks like — paste 1–3 samples).
Compare a weak prompt — "Write a LinkedIn post about AI" — to a strong one that specifies role (a growth marketer), context (writing for B2B SaaS founders), task (argue companies buy AI before fixing process), format (structure, word count, banned words), and an example post to match in style. The strong prompt takes 90 seconds to write and saves 20 minutes of editing.
Context is the highest-leverage ingredient
Of the five components, context moves quality the most. Modern models can infer role and format reasonably well, but they cannot invent facts about your business. Practical ways to inject context: paste it directly (your website copy, proposals, meeting notes handle long pastes well); upload files (PDFs, spreadsheets, decks — both ChatGPT and Claude read them directly); use Projects, which let you create standing instructions and reference files that apply to every chat inside them, so you stop re-explaining your business every time; and let the model ask — ending complex prompts with "before you start, ask me up to 5 questions that would meaningfully improve the output" routinely doubles output quality.
Iteration — the second prompt matters more than the first
Amateurs treat the first output as the final answer. Operators treat it as a first draft from a junior colleague. Your follow-up vocabulary: **Tighten** ("cut this by 40% without losing the argument"), **Redirect** ("good structure, wrong tone — make it warmer"), **Deepen** ("expand point 2 with a concrete example"), **Variant** ("give me 3 alternative hooks, ranked"), **Critique** ("act as a skeptical customer, tear this apart, then rewrite addressing your own critique"), and **Extract** ("turn this into a reusable template with placeholders").
That last move is important: whenever a conversation produces something great, spend one more prompt converting it into a reusable template or checklist. That's how one-off wins become a system.
Advanced moves worth knowing
**Meta-prompting**: ask the AI to write the prompt — "I want [outcome]. Write the ideal prompt I should give you, ask me for missing information, then execute it." This works because the model knows what information it needs better than you do.
**Chain-of-thought for hard problems**: for analysis, planning, or math, say "think through this step by step" or switch on extended thinking mode. **Structured output**: ask for a table or valid JSON when the output needs to be machine-usable — essential for automation pipelines. **Negative constraints**: models respond well to explicit bans, like "do not use bullet points" or "do not invent statistics." **Style transplants**: paste 2–3 samples of your writing and ask the model to describe your style in 10 bullets, then apply those rules to everything it writes.
A prompt library beats prompt genius
You don't need to be creative at 9am on a Tuesday. You need a document with your 10–20 proven prompts, ready to paste. The course's Appendix A gives a starter library of 25 prompts across email, research, content, analysis, and planning — treat it as seed stock and evolve your own. Store your library where you actually work: a Notion page, a text file, or directly inside Project instructions so you never have to paste them at all.
How to run it
- Captured a lazy baseline. You ran your real task as a one-line prompt first, without embellishment, to have something honest to compare against.
- Built a full five-component prompt. Your rewritten prompt includes role, context, task, format, and an example, using real details from your situation.
- Compared outputs concretely. You wrote specific differences (not just "better") between the lazy and full-brief outputs.
- Applied one iteration move. You used exactly one named verb (tighten, redirect, deepen, variant, critique) and observed its distinct effect.
- Extracted a reusable template. You converted the winning prompt into a template with placeholders and saved it to a findable prompt library.
In the field
A consultant needs a proposal follow-up email. His lazy prompt — "write a follow-up email" — produces a generic, forgettable draft. His five-component rewrite specifies role (senior consultant), context (the client went quiet after a $40k proposal three weeks ago), task (a warm but direct nudge that offers a smaller pilot option), format (under 120 words, no corporate jargon), and an example of his own past successful follow-up. The result reads like him, not like a template — and after one "tighten" pass, it's ready to send.
When not to use it
The full five-component brief is overkill for truly trivial, disposable requests — reformatting a sentence, a quick synonym check, or a one-off joke doesn't need role, context, and examples spelled out. Reserve the full brief for outputs you'll actually use or send; for genuinely throwaway micro-tasks, a lazy one-line prompt is the right level of effort and insisting on a full brief every time just adds friction without adding value.
Pitfalls and takeaways
- Writing a one-line lazy prompt and treating the first output as final instead of a draft
- Forgetting to supply context — the model can't invent facts about your business
- Re-explaining your business from scratch in every new chat instead of using Projects
- Asking for creative or unstructured output when the result needs to feed into an automation (should be a table or JSON)
- Not asking the model to critique or argue against its own output
- Prompt = brief: role, context, task, format, examples
- Context is the highest-leverage ingredient; Projects let you inject it permanently
- The first output is a draft — iteration verbs (tighten, redirect, deepen, critique) do the real work
- "Ask me 5 questions before you start" and meta-prompting are the two best power moves
- Build a prompt library; systems beat inspiration
Do the work
Goal: Rewrite a real lazy prompt using all five brief components, run one iteration move on the result, and extract it into a reusable template — building the first entry in your prompt library. (25 min)
- Write and run the lazy version. In a fresh chat, type a one-line prompt for your real task exactly as you normally would (e.g., "write a follow-up email to a client"). Run it and read the output without editing.
- Rebuild the prompt with all five components. In a new chat, rewrite the prompt with role, context, task, format, and example explicitly included, using real details from your actual situation.
- Compare the two outputs side by side. Read both outputs back to back and write two sentences on what specifically changed — tone, specificity, usefulness.
- Run one iteration move. On the stronger output, apply exactly one iteration verb from the module (tighten, redirect, deepen, variant, or critique) and note how the result changes.
- Extract a reusable template. Ask the model: "Turn this final prompt into a reusable template with placeholders I can reuse for similar tasks." Save the resulting template.
Deliverable: A saved prompt-library entry containing: the original lazy prompt, the five-component rewrite, the iteration move applied, and the final reusable template with placeholders.
Sources
- OpenAI, "Prompt Engineering Guide," platform.openai.com/docs/guides/prompt-engineering
- Anthropic, "Prompt Engineering Overview," docs.anthropic.com
- White et al., "A Prompt Pattern Catalog to Enhance Prompt Engineering with ChatGPT," arXiv, 2023
- Google, "Prompting Guide 101," Google Workspace AI resources