ChatGPT prompts for work: 70 templates for common tasks

ChatGPT prompts for work produce better results when you define role, task, context, constraints, and output format before you ask for the draft.

What prompt settings for work should you define before using templates?
Which 70 prompts for work cover the most common tasks?
How should you verify each prompt output before sending or publishing it?
Which prompt mistakes for work should you avoid to reduce risk?
How can you adapt these templates to your team so they stay consistent?

What prompt settings for work should you define before using templates?

Prompt settings for work should be defined first, so the response fits your workflow on the first pass.

Google, in its Google Workspace prompting guide, recommends structuring prompts around four anchors: role, task, context, and format. That structure alone usually reduces rework.

Use this reusable base frame for any work task:

Role: who the model should act as
Goal: what outcome you need
Context: facts, limits, audience
Format: list, email, table, plan
Check: what makes the answer acceptable

Validate the first output for accuracy, workflow fit, and unsupported assumptions. If it misses, add one constraint or one format rule instead of rewriting the whole prompt.

Which 70 prompts for work cover the most common tasks?

These 70 prompts for work cover messaging, meetings, summaries, planning, analysis, writing, support, and internal operations.

Which prompts for email and short messages are useful every day?

Prompts for email and short messages work best when you specify recipient, tone, and next action.

Client email: [topic], [position], [next step].
Complaint reply: [issue], [resolution], [timeline].
Team reminder: [task], [deadline], [owner].
Manager update: [status], [risk], [need].
Delay apology: [reason], [new date].
Meeting confirmation: [time], [goal], [agenda].
Polite decline: [request], [reason], [alternative].
Clarification request: [missing info], [what needed].
Post-call email: [summary], [actions], [dates].
Chat message: [context], [action], [deadline].

Which prompts for meetings and notes reduce missed details?

Prompts for meetings and notes reduce missed details when you ask separately for decisions, actions, and open questions.

Agenda draft: [goal], [participants], [time blocks].
Prep questions: [topic], [decision], [risks].
Meeting notes: [raw notes], [decisions], [owners].
Leadership summary: [decided], [next steps].
Action list: [action], [owner], [due date].
Open questions list: [items], [who to assign].
Meeting risks: [risk], [likelihood], [mitigation].
Options decision: [options], [criteria], [recommendation].
Update for absentees: [summary], [read next].
Stand-up format: [done], [next], [blockers].

Which prompts for summaries and explanations save review time?

Prompts for summaries and explanations save review time when you set length and audience before the model writes.

Summary of text: [text], [5 bullets].
Thread summary: [email thread], [decisions], [actions].
Executive summary: [document], [1 paragraph].
Explain simply: [topic], [audience].
Extract key risks: [text], [list].
Extract deadlines: [text], [dates], [owners].
Notes to action plan: [draft notes], [steps].
Compare two versions: [v1], [v2].
Decisions from document: [text], [table].
Team chat summary: [text], [short format].

Which prompts for planning and prioritization help move work forward?

Prompts for planning and prioritization help move work forward when you define criteria and resource limits.

Daily plan: [tasks], [time], [priority].
Weekly plan: [goals], [constraints], [checkpoints].
Prioritize tasks: [list], [criteria], [order].
Break down project: [goal], [phases], [outputs].
Plan risk review: [plan], [risks], [fallbacks].
Launch plan: [date], [roles], [dependencies].
Meeting prep plan: [goal], [materials], [questions].
Onboarding plan: [role], [30 days].
Handover plan: [items], [owner], [timing].
QA check plan: [stages], [criteria].

Which prompts for analysis and decisions produce structured answers?

Prompts for analysis and decisions produce structured answers when you request criteria, options, and limits on the conclusion.

Compare options: [A], [B], [criteria].
Pros and cons: [decision], [context].
Root-cause analysis: [symptom], [data].
Testable hypotheses: [problem], [signals].
Recommendation by constraints: [constraints], [options].
Change-risk analysis: [change], [impact], [controls].
Scenario planning: [base], [best], [worst].
Vendor criteria: [need], [weights].
Assumption-check questions: [plan], [risks].
Manager recommendation: [analysis], [decision], [conditions].

Which prompts for writing and editing improve clarity without over-styling?

Prompts for writing and editing improve clarity when you request a specific edit type instead of a generic improvement.

Rewrite clearly: [text], [audience], [tone].
Shorten without loss: [text], [target length].
Make more formal: [text], [level].
Make more tactful: [text], [limits].
Check argument logic: [text], [weak points].
Convert to bullets: [text], [count].
Bullet points to email: [points], [recipient].
Text to table: [text], [columns].
Ambiguity review: [text], [risk areas].
Draft 3 headings: [topic], [tone].

Which prompts for support, sales, and internal operations should stay ready?

Prompts for support, sales, and internal operations should stay ready because those tasks repeat across teams.

Support reply: [request], [policy], [steps].
Escalation template: [issue], [impact], [urgency].
Client explanation: [complex topic], [plain language].
Qualification questions: [customer type], [goal].
Sales call summary: [notes], [next step].
Internal procedure draft: [process], [steps], [errors].
FAQ draft: [topic], [10 items].
Status report template: [project], [progress], [risks].
Resource request template: [need], [reason], [date].
Retrospective template: [worked], [change next].

How should you verify each prompt output before sending or publishing it?

Each prompt output should be verified with the same short checklist, so mistakes do not flow into emails, documents, or decisions.

NIST AI RMF 1.0 emphasizes characteristics such as validity and reliability, which makes a repeatable verification step especially useful for work outputs.

Check facts, numbers, names, and dates.
Check whether the format matches the task.
Check for invented details or overreach.
Check for a clear next action.

If one check fails, send one corrective line: what is wrong, what to fix, and which format must stay unchanged.

Which prompt mistakes for work should you avoid to reduce risk?

Prompt mistakes for work should be avoided first in tasks that involve confidential data, access details, or internal records.

NIST IR 8446 notes that generative AI introduces additional cybersecurity risks, so do not paste passwords, tokens, client personal data, or raw internal tables into prompts.

Common mistakes:

one prompt tries to do multiple tasks
no output format is defined
no quality criteria are set
sensitive data is pasted
no verification step before use

If the task is sensitive, use anonymized inputs and ask for a structure or template first, then fill real values yourself.

How can you adapt these templates to your team so they stay consistent?

These templates stay more consistent when you add 1-2 examples of a good result and lock the response format for repeated tasks.

Google also recommends examples and explicit formatting, so build a small internal set of sample outputs for emails, summaries, reports, and meeting notes.

Start with 5 templates your team uses every week, align the expected output format, then expand to 15-20. That turns a prompt list into a practical working standard.

Sources: