PromptBro

ChatGPT Prompt Templates — Copy, Paste, and Adapt

A good ChatGPT prompt template isn't just a fill-in-the-blanks form — it's a structural pattern that ensures you never forget the elements that make prompts work. Below are five reusable template frameworks for common tasks, each shown with a real filled-in example so you can see exactly how the structure applies. Copy the template, adapt the bracketed sections, and get significantly better results than open-ended requests.

The anatomy of a ChatGPT prompt

Every high-performing ChatGPT prompt contains five structural elements. Most bad prompts are missing at least two of them:

  • Role — "You are a [expert type]..." Sets the persona, expertise level, and perspective.
  • Context — Background on your situation: who you are, what you're working on, what you know, what you've tried.
  • Task — The specific, action-verb instruction: write, analyze, summarize, compare, generate, debug.
  • Format — How you want the output structured: bullet list, numbered steps, table, prose, JSON, code block.
  • Constraints — Length limits, tone requirements, what to avoid, reading level, language style.

5 Reusable Prompt Templates

Template 1 — Writing & content

You are a [EXPERT ROLE, e.g. "copywriter specializing in B2B SaaS"].

Context: [YOUR SITUATION, e.g. "I'm writing for a product that helps HR teams automate onboarding.
My audience is HR managers at mid-size companies who are skeptical of new software."]

Task: Write a [FORMAT, e.g. "500-word blog post"] titled "[TITLE]".
Structure: [e.g. "hook paragraph, 3 numbered points each with a real-world example, closing CTA"]
Tone: [e.g. "conversational but credible. Not salesy."]
Avoid: [e.g. "jargon, passive voice, phrases like 'in today's fast-paced world'"]

— FILLED EXAMPLE —

You are a B2B copywriter specializing in HR tech.

Context: I'm writing for an onboarding automation platform. Audience: HR managers at 200–500
person companies who have tried and abandoned manual onboarding checklists.

Task: Write a 500-word blog post titled "Why Your Onboarding Process Is Losing You the Employees You Just Hired."
Structure: emotional hook (2 sentences), problem framing (100 words), 3 specific failure modes
(each 80 words with a real-world scenario), solution hint (50 words), CTA.
Tone: empathetic but direct. Not salesy.
Avoid: buzzwords, passive voice, vague advice.

Template 2 — Analysis & research

You are a [EXPERT ROLE, e.g. "market research analyst with expertise in SaaS"].

I need an analysis of [TOPIC].
Context: [WHY YOU NEED THIS, e.g. "I'm deciding whether to enter the project management software
market with a niche product for law firms."]

Analyze:
1. [FIRST DIMENSION, e.g. "Market size and growth trajectory"]
2. [SECOND DIMENSION, e.g. "Key competitors and their positioning"]
3. [THIRD DIMENSION, e.g. "Underserved segments or whitespace"]

Format: [e.g. "structured memo with headers. 600 words max. Lead each section with the key finding."]
Be direct. Flag where you're inferring from limited data vs. citing known facts.

— FILLED EXAMPLE —

You are a market research analyst specializing in legal tech and SaaS.

I need an analysis of the project management software market for law firms.
Context: I'm evaluating whether to build a niche PM tool specifically for small litigation teams.

Analyze:
1. Current state of law firm project management (tools in use, satisfaction levels)
2. Why general PM tools (Asana, Monday) don't fully serve law firms
3. Whitespace: what a purpose-built tool would need to win

Format: structured memo, 600 words max. Lead each section with the key finding.
Flag where data is inferred vs. cited.

Template 3 — Code & technical

You are a [EXPERT ROLE, e.g. "senior Python engineer with expertise in data pipelines"].

Task: [SPECIFIC TECHNICAL TASK, e.g. "Write a function that..."]
Requirements:
- [REQUIREMENT 1]
- [REQUIREMENT 2]
- [REQUIREMENT 3]

Constraints:
- [e.g. "Use only standard library — no third-party packages"]
- [e.g. "Include type hints and a docstring"]
- [e.g. "Handle edge cases: empty input, None values, malformed data"]

Format: code block first, then a brief explanation of key decisions. No unnecessary prose.

— FILLED EXAMPLE —

You are a senior Python engineer who prioritizes readable, maintainable code.

Task: Write a function that parses a CSV file of transactions and returns a summary dict
with total spend per category, filtered to a date range.
Requirements:
- Accept file path, start date, and end date as parameters
- Handle missing or malformed rows gracefully (log a warning, skip the row)
- Return results sorted by spend descending

Constraints:
- Standard library only (csv, datetime, logging)
- Full type hints and docstring
- Handle edge cases: empty file, no rows in date range, missing category column

Format: code block first, then 3–5 bullet explanation of key design choices.

Template 4 — Creative work

You are a [CREATIVE ROLE, e.g. "literary fiction author in the tradition of [Author]"].

Write a [FORMAT, e.g. "600-word short story opening scene"] for a [GENRE] story.
Premise: [CORE SETUP, e.g. "A forensic accountant discovers her firm has been laundering
money for a client she personally likes."]

Tone: [e.g. "quiet dread, not thriller-paced — let tension build slowly"]
POV: [e.g. "close third person, present tense"]
Style constraints: [e.g. "Show don't tell. Grounded in specific sensory detail.
Avoid melodrama. The protagonist should feel real, not like a genre archetype."]
End: [e.g. "on an ambiguous moment that creates forward momentum without a cliffhanger cliché"]

— FILLED EXAMPLE —

You are a literary crime fiction author in the tradition of Tana French and Celeste Ng.

Write a 600-word opening scene for a psychological thriller.
Premise: A forensic accountant at a mid-size firm discovers her most likable client has been
using the firm's services to launder money from a source she can't yet identify.

Tone: quiet, creeping dread. Not thriller-paced.
POV: Close third person, past tense.
Style: Grounded in the mundane details of office life. The wrongness should feel like something
noticed peripherally, not dramatically. No internal monologue spelling out what she suspects.
End: on a small, specific detail that a reader could miss but shouldn't.

Template 5 — Planning & strategy

You are a [EXPERT ROLE, e.g. "product strategy consultant"].

Context: [YOUR SITUATION]
Goal: [WHAT YOU'RE PLANNING, e.g. "Launch a new product feature to existing customers in Q3."]
Constraints: [KEY LIMITS, e.g. "2 engineers, $5K marketing budget, 8 weeks until launch"]

Build a [DELIVERABLE TYPE, e.g. "launch plan"] covering:
1. [SECTION 1, e.g. "Pre-launch: customer validation and messaging"]
2. [SECTION 2, e.g. "Launch week: channels and sequencing"]
3. [SECTION 3, e.g. "Post-launch: success metrics and iteration triggers"]

Format: structured plan with headers. Be specific — timelines, owners, concrete actions.
Skip the obvious. I need tactical clarity, not strategic platitudes.

— FILLED EXAMPLE —

You are a product strategy consultant who has launched 10+ SaaS features to existing customer bases.

Context: I run a time-tracking SaaS with 2,400 paying customers. We're adding a budget forecasting
feature built on top of tracked hours.
Goal: Successful launch to existing customers with at least 20% activation in 60 days.
Constraints: 2 engineers (post-launch support only), $5K marketing budget, 8 weeks.

Build a launch plan covering:
1. Beta phase (weeks 1–4): who to invite, how to collect feedback, what "done" looks like
2. Full launch (weeks 5–6): email sequences, in-app prompts, success metric targets
3. Post-launch (weeks 7–8): activation review, iteration triggers, churn-risk monitoring

Format: headers, sub-bullets, specific timelines. No filler.

Tips for adapting templates

Don't just swap out the bracketed placeholders mechanically. Read the whole template and think about whether the structure fits your actual goal. If you're writing a short Slack message, you don't need the full five-element structure. If you're writing a complex analysis for a board presentation, you might need more sections. Templates are patterns, not rules.

The "Avoid" field is underused and high-value. ChatGPT has strong defaults that can override even good prompts — things like excessive hedging, generic examples, or filler phrases like "it's important to note." An explicit "Avoid" list cuts those defaults off before they appear.

If the output is good but not quite right, don't rewrite the whole prompt. Follow up in the same conversation: "Make the tone slightly more formal," or "Cut the third section — it's not relevant to my use case." ChatGPT handles in-conversation refinement very well.

Why use PromptBro?

Templates are useful, but they require you to know what to fill in. PromptBro figures that out for you. Describe your goal in plain language, answer a few guided questions, and get a complete, filled-in prompt calibrated to your situation — no manual template adaptation required. It's faster than editing a template and produces better output because every field is informed by your specific context.

Try PromptBro free — build your first prompt in 60 seconds →

Related resources