5 ChatGPT Prompts Everyone Should Know

Everyone who uses ChatGPT should know these 5 types of prompt!

They’re your Swiss-Army knife – the multi-tool of ChatGPT prompts that you’ll reach for again and again.

Learning these prompt structures is much more valuable than learning individual prompts. Because these templates below are so versatile they can be applied to almost anything in your personal life or your professional work. You will create genuinely useful ChatGPT prompts that will get you deep insights and good quality outputs.

So here they are with real-world examples.

1. Flipped Interaction Prompt

• Problem it solves – long, bloated prompts that still miss the mark.

• How it works – you set the goal, then force ChatGPT to ask you questions one at a time until it can solve your problem. If it’s a work task, you can also give it a professional role, like lawyer or marketer.

• Skeleton — You are a [role]. Goal: [objective]. Ask me questions one-by-one until you can solve it. Then deliver the answer.

• Example — You’re a workflow analyst. Goal: save me five hours a week on admin. Ask me questions one at a time until you can achieve that for me.

• ChatGPT fires back with focused questions; after 3–5 rounds it returns an ordered list of fixes with first steps.

2. Role-Based Prompt

• Problem it solves – generic, middle-of-the-road advice.

• How it works – lock the model into a specific professional persona, so tone, vocabulary, and priorities match the role.

• Skeleton — You are a senior [profession]. Perform [task] on the following text/data.

• Example — You are a Senior Project Manager with deep expertise in construction. Look through this project plan in detail and list specific problems, areas for improvement, and missed opportunities.

• Pro tip – add seniority and years of experience for extra focus.

3. Self-Reflection Loop

• Problem it solves – more accurate, less biased, higher-quality outputs.

• How it works – force ChatGPT to produce a draft, critique it, then rewrite. There’s evidence that this improves results. Particularly great for writing.

• Skeleton — Write [output brief]. In ≤ [word limit] critique your draft for clarity, jargon, and persuasive strength. Rewrite once, fixing every issue. Return the improved version and one sentence on the biggest change.

• Example — Write a 150-word LinkedIn post that encourages small-business owners to adopt email-automation tools. Next, in ≤ 100 words, critique your draft for clarity, jargon, and persuasive strength. Finally, rewrite the post once, fixing every issue you flagged. Return the improved post, then one sentence summarising the main change you made.

• Watch-out – keep word limits tight; long reflections get rambling.

4. Forced Ranking Scorecard

• Problem it solves – woolly, poorly thought-through comparisons.

• How it works – make the model assign numbers to each criterion, then sort by total. Quantification forces concrete thinking.

• Skeleton — List [n] options. Score each 0-10 on [criteria]. Show a table, include totals, sort high→low. Justify the winner in one short paragraph.

• Example — Find five trail-running shoes; score cost, stability, comfort, durability from 1-10 in a table. Choose the winner and write a paragraph justifying it as the best choice. You get a ranked table and a clear winner with caveats.

• Pro tip – weight criteria (e.g., cost × 0.5) if some matter more.

5. 5 Whys Root-Cause Drill-Down

• Problem it solves – treating symptoms instead of causes.

• How it works – ChatGPT asks “Why?” up to five times, each answer feeding the next question. Technique comes from Toyota’s Taiichi Ohno.

• Skeleton — Problem: [statement]. Ask “Why?” up to five times. Summarise the root cause in one sentence. Give one fix I can apply this week.

• Example — Open rate dropped from 38 % to 24 %. Ask why until root cause is clear. The model might trace the drop to a new sender domain that lands in spam, then recommend a domain warm-up plan.

• Watch-out – stop early if the root cause emerges in fewer than five whys.