120 Useful ChatGPT Prompt Patterns for Work

ChatGPT prompts do not instantly boost productivity by themselves. They work when they help you give better instructions: clear task, useful context, audience, output format, and constraints.

OpenAI’s prompt guidance emphasizes being clear and specific, providing context, defining format, and refining iteratively. Source: OpenAI ChatGPT prompt best practices.

Use the patterns below as templates, not scripts.

ChatGPT also changes over time. OpenAI’s release notes describe ongoing updates such as Projects improvements, study mode, record mode, model support changes, and personalization updates. That means the best prompt is not a magic sentence. The best prompt is a clear working brief that can travel across models, tools, and features.

The Core Work Prompt

Help me with [task].

Context:
[background, audience, goal, constraints]

Output:
[format, length, tone, structure]

Quality bar:
[what a good answer must include or avoid]

Ask clarifying questions only if needed. Otherwise, proceed with reasonable assumptions.

How to Customize the Prompts

For better results, add five ingredients:

1. Role: who ChatGPT should act as.
2. Context: what it needs to know.
3. Task: the work to perform.
4. Format: the shape of the output.
5. Quality bar: what would make the answer useful.

Example:

Act as a product marketing manager.
Context: We sell project management software to agencies with 10-100 employees.
Task: Draft a comparison page outline against spreadsheets.
Format: H1, intro, comparison table, objection section, FAQ, CTA.
Quality bar: Be specific, avoid fake statistics, and identify claims that need proof.

This is stronger than “write a comparison page” because it gives the model a job, audience, structure, and guardrails.

Writing and Content

1. Draft a [document type] for [audience] about [topic]. Include [must-cover points] and avoid [things to avoid].
2. Turn this outline into a first draft: [outline]. Keep the tone [tone] and the length about [length].
3. Edit this for clarity without changing meaning: [text].
4. Rewrite this for [audience] who cares about [priority].
5. Create 10 headline options for [content], grouped by benefit, curiosity, and directness.
6. Summarize this into [number] bullets for [audience]: [text].
7. Identify unsupported claims in this draft and suggest safer wording.
8. Make this more scannable with headings, shorter paragraphs, and bullets.
9. Turn this blog post into a LinkedIn post series.
10. Write three introductions for this topic using different hooks.

Project Management

11. Break this project into tasks, dependencies, and milestones: [project].
12. Create a project brief with goals, scope, stakeholders, timeline, and risks.
13. Build a meeting agenda for [meeting] with [duration] and [participants].
14. Turn these meeting notes into decisions, action items, owners, and deadlines.
15. Identify the top risks in this plan and suggest mitigations.
16. Create a status update from these notes: [notes].
17. Define success criteria for [initiative].
18. Create a retrospective framework for [project].
19. Map dependencies and critical path for [project].
20. Draft stakeholder communication for [change].

Marketing and Sales

21. Create a campaign brief for [offer] targeting [audience].
22. Draft a landing page structure for [product] with headline, proof, objections, and CTA.
23. Write an email sequence for [audience] after [trigger].
24. Build a customer persona from this data: [data].
25. Compare our positioning with competitors: [competitors].
26. Generate A/B test ideas for [page/email/ad].
27. Create a content calendar for [platform] around [theme].
28. Draft sales call prep for [prospect] using this context: [context].
29. Write objection-handling responses for [objections].
30. Turn customer feedback into messaging themes.

Research and Analysis

31. Research [topic] using current, primary sources and return a source-linked briefing.
32. Compare [option A] vs [option B] across [criteria].
33. Analyze this data and identify patterns, anomalies, and questions: [data].
34. Create a decision memo for [decision].
35. Identify what evidence would support or refute this claim: [claim].
36. Summarize the strongest arguments on both sides of [debate].
37. Find assumptions in this analysis: [analysis].
38. Create a market overview for [category] with risks and open questions.
39. Turn these sources into a synthesis, noting agreement and disagreement.
40. Build a checklist to verify claims in this draft.

Strategy and Planning

41. Develop three strategic options for [situation].
42. Run a SWOT analysis for [company/product].
43. Create OKRs for [team] this quarter.
44. Build a 30-day execution plan for [goal].
45. Prioritize these initiatives using impact, effort, risk, and urgency: [list].
46. Create a business case for [proposal].
47. Build a scenario plan around [uncertainty].
48. Identify the highest-leverage activities for [goal].
49. Create a resource plan for [project].
50. Stress-test this plan against pessimistic, likely, and optimistic cases.

Personal Productivity

51. Help me plan today around these priorities and meetings: [list].
52. Create a weekly review from these notes: [notes].
53. Identify what I should delegate from this workload: [workload].
54. Design a focus block schedule for [type of work].
55. Create an email triage system for [role/context].
56. Help me say no to [request] while preserving the relationship.
57. Build a learning plan for [skill] with [hours/week].
58. Create a decision journal entry for [decision].
59. Design a shutdown routine for the end of my workday.
60. Help me identify time sinks in this week: [schedule].

Coding and Technical Work

61. Review this code for bugs, edge cases, and readability: [code].
62. Explain this code to a [beginner/intermediate/expert]: [code].
63. Generate tests for this function: [code].
64. Refactor this for clarity without changing behavior.
65. Write technical documentation for [feature/API].
66. Debug this error with likely causes and next checks: [error].
67. Compare implementation approaches for [feature].
68. Create a migration plan for [system change].
69. Turn this user story into technical tasks.
70. Review this API design for consistency and failure cases.

Communication

71. Draft a difficult conversation opener for [situation].
72. Write feedback that is specific, kind, and actionable.
73. Prepare talking points for [meeting/presentation].
74. Rewrite this message to be clearer and less defensive.
75. Draft a customer apology that takes responsibility and explains the fix.
76. Create a concise executive update from [details].
77. Turn this technical explanation into plain language.
78. Prepare questions for [interview/meeting].
79. Write a follow-up email after [event].
80. Summarize this disagreement and propose a fair next step.

Creative Work

81. Generate 20 ideas for [creative challenge] with different angles.
82. Create names for [product/project] that feel [qualities].
83. Develop a mood board description for [concept].
84. Write five story hooks for [premise].
85. Create variations of this idea in different styles: [idea].
86. Identify what is generic in this creative concept.
87. Suggest a stronger visual metaphor for [message].
88. Turn this rough idea into a pitch.
89. Create a creative brief for [campaign/project].
90. Find the emotional core of this concept: [concept].

Review and Quality Control

91. Check this draft for factual claims that need sources.
92. Review this plan for risks, assumptions, and missing owners.
93. Critique this proposal like a skeptical executive.
94. Identify ambiguity in this policy draft.
95. Review this content for tone, clarity, and trustworthiness.
96. Find compliance-sensitive statements in this copy.
97. Check whether this answer actually addresses the question.
98. Create a pre-publish checklist for [content].
99. Compare the final draft against the brief.
100. Suggest edits that improve credibility without adding hype.

Better Follow-Ups

101. Make this more specific.
102. Shorten this by 30 percent without losing meaning.
103. Give me three stronger alternatives.
104. Challenge your own answer.
105. What would an expert disagree with here?
106. What assumptions are you making?
107. Turn this into a checklist.
108. Add examples.
109. Remove unsupported claims.
110. Make it more practical.
111. Rewrite for a skeptical reader.
112. Explain the tradeoffs.
113. Provide sources for the factual claims.
114. Separate facts from recommendations.
115. Identify the riskiest part.
116. Make the tone more human.
117. Make it less promotional.
118. Convert this to a table.
119. Add next steps.
120. Summarize what changed.

Prompt Safety Rules for Work

Prompting well is not only about getting a clever answer. It is also about avoiding bad outputs.

Use these rules:

• Do not paste confidential data unless your company allows it.
• Ask for sources when factual accuracy matters.
• Check legal, medical, financial, HR, and compliance content manually.
• Do not ask ChatGPT to invent proof, quotes, citations, or case studies.
• Use human review before publishing customer-facing work.
• Keep prompts specific enough that the output can be checked.
• Separate brainstorming from final decisions.

For business use, the safest workflow is draft, verify, edit, approve.

How to Turn Prompts Into Repeatable Workflows

If a prompt works more than twice, turn it into a reusable workflow:

1. Save the prompt pattern.
2. Add required inputs.
3. Add review steps.
4. Add examples of good output.
5. Add failure cases to avoid.
6. Decide who owns updates.

This is especially useful for recurring work such as weekly reports, meeting summaries, sales follow-ups, SEO briefs, support replies, hiring scorecards, and project status updates.

For teams, consider storing prompts in a shared document or project workspace. The value is not the prompt alone. The value is a shared way of working.

Example Prompt Workflows

Meeting summary workflow:

Turn these notes into:
1. Decisions
2. Open questions
3. Action items with owners
4. Risks
5. A short follow-up email
Notes: [paste notes]

Research workflow:

Create a source-linked briefing on [topic].
Use current primary sources where possible.
Separate facts, interpretation, and recommendations.
Flag anything uncertain.

Editing workflow:

Edit this draft for clarity, credibility, and usefulness.
Do not change the meaning.
Remove hype.
Mark any claim that needs evidence.
Draft: [text]

Prompt Patterns by Skill Level

Beginner prompts should focus on structure. Ask ChatGPT to organize information, create outlines, summarize notes, and draft simple first versions. The goal is to reduce blank-page friction.

Intermediate prompts should include constraints. Add audience, tone, examples, source requirements, word count, format, and review criteria. The goal is to get output that fits a real workflow.

Advanced prompts should include iteration. Ask ChatGPT to produce a draft, critique it, identify assumptions, improve weak sections, and produce a final version with changes explained. The goal is not one-shot output. The goal is a better thinking partner.

Common Prompt Mistakes

The first mistake is asking for too much at once. If you ask for strategy, copy, SEO, sources, design, analytics, and implementation in one prompt, the output often becomes shallow. Break complex work into steps.

The second mistake is hiding the audience. A prompt for executives should produce a different answer than a prompt for new employees, engineers, customers, or students.

The third mistake is accepting the first answer. Good prompting is conversational. Ask for alternatives, ask what is missing, and ask the model to critique its own assumptions.

The fourth mistake is treating generated text as evidence. ChatGPT can help you find what needs evidence, but facts still need verification.

Final Practical Rule

Use ChatGPT like a capable colleague who needs context. Give it the brief you would give a human. Then review the work like you would review a human draft.

That mindset produces better prompts than any memorized list.

References

• OpenAI Help: Prompt engineering best practices for ChatGPT
• OpenAI Help: ChatGPT release notes
• OpenAI Help: Custom instructions for ChatGPT

Bottom Line

The point is not to memorize 120 prompts. The point is to learn the pattern: task, context, output, quality bar, and iteration. Once you understand that structure, you can make better prompts for almost any kind of work.