The best AI prompts for executives, 22 to copy (2026)

The best AI prompts for executives give the tool a role, the context of your department, a specific task and an output format. Here are 22 ready to use in ChatGPT, Claude, Copilot or Gemini. If you want a complete path, see the AI program for executives.

Add Founderz as a preferred source

What you’ll take from this

This library gathers the best AI prompts for executives, effective prompts for ChatGPT, Claude, Copilot and Gemini built for everyday repetitive tasks.

• 22 AI prompts for executives ready to copy: data analysis, project management, email, meetings and marketing.
• Which tool fits which task: ChatGPT for open-ended analysis, Claude for long documents, Copilot inside Word, Excel, Outlook and Teams, Gemini with Google Workspace.
• The five fields of an effective prompt and an example of a weak prompt turned into a usable one.
• A method for spotting bottlenecks in your team’s workflow with data you already have.
• How to build a prompt library your leadership team actually uses, and what data should never go into any tool.

If you run a department or a team, you already know where the time goes: long reports to summarize, the same email every week, follow-ups nobody enjoys writing. These prompts target that layer of work: they help you analyze data, design plans, evaluate options, review documents and propose measurable actions. They do not replace your judgment, but they claw back hours of repetitive tasks and improve the quality of the response you bring to every meeting.

ChatGPT prompts for executives

ChatGPT works well for open-ended analysis, document synthesis and first drafts. Swap what is in brackets for your own details and keep the rest: each prompt follows the structure of role, context, task, constraints and format.

Data analysis and decision-making prompts

Quarterly report summary for the leadership committee

Act as a business analyst with experience in reporting for executive committees.
Context: I am sending you the sales report for last quarter from [company or department name]. [Paste report or data]
Task: summarize it in five points for the committee. Highlight deviations from the target of [figure or percentage] and flag which data needs more investigation before it is presented.
Constraints: do not invent figures that are not in the report. Mark what comes from the document as [DATA] and what you infer as [INTERPRETATION].
Format: one bold headline per point and one supporting sentence.

Compare two hiring options

Act as a strategy consultant with experience in sales teams.
Context: I lead the sales department at [a company in industry X], with [N] people. I want to grow the team and I am torn between [three junior profiles] and [two senior profiles]. Annual budget: [amount].
Task: compare both options on cost, time to productivity and risk. Add a recommendation and the conditions under which it would stop being valid.
Constraints: use only the data I gave you and ask me before assuming anything else.
Format: comparison table, then the recommendation in one paragraph.

Hypotheses about employee turnover

Act as a people analytics specialist.
Context: I am sending you the turnover data for the last six months at [company], by department, tenure and reason for leaving where recorded. [Paste data]
Task: identify patterns by department and propose three hypotheses about the causes, each with the evidence behind it and the data I am missing to confirm it.
Constraints: do not attribute causes to specific people and clearly mark what is data and what is interpretation.
Format: three blocks with hypothesis title, evidence and next check.

AI-assisted project management and tracking prompts

Launch plan by phases

Act as a project manager with experience in product launches.
Context: I am launching [a new feature of my product] in [three months]. Team: [N people from product, engineering and marketing]. Known dependencies: [vendor, legal, another department].
Task: break the project into phases, with milestones and dependencies between tasks, and point out the critical path.
Constraints: assign tasks to roles, not to named people. If information is missing, ask three questions first.
Format: week-by-week timeline as a table with phase, milestone, owner (role) and dependency.

The five most likely risks in a plan

Act as a project risk manager.
Context: this is the plan for [project] with a delivery date of [date] and a budget of [amount]. [Paste plan]
Task: identify the five most likely risks. For each one give the likelihood (high, medium or low), the impact and a concrete mitigation action with an owner.
Constraints: prioritize risks specific to this plan, not generic risks of any project.
Format: table ordered from highest to lowest exposure.

Questions for the weekly check-in

Act as a leadership team coach.
Context: I run the weekly check-in with my team of [N people] in [department]. This week the project [name] is [one week] behind.
Task: prepare six questions that cover progress, blockers and next week's priorities, and end with a question that closes on a decision.
Constraints: direct, non-punitive tone, with no questions that look for someone to blame.
Format: numbered list with suggested time per question, 30 minutes in total.

Claude prompts for executives

Claude stands out with long documents and careful drafting of sensitive texts. Use it when you have to read a lot before deciding or when tone matters. If your horizon is the top job, see also the AI prompts for CEOs.

Long documents and sensitive communication

From a 20-page technical document to a committee summary

Act as an adviser to the leadership committee.
Context: I am attaching a technical document of [20 pages] about [topic, for example the migration of the billing system]. The committee is not technical and has [five minutes] to decide.
Task: summarize the document in 200 words, with the three conclusions first, the decision being requested and the main risk.
Constraints: no jargon, do not repeat the document, and warn me if any conclusion is not supported by what the text says.
Format: three numbered conclusions, one paragraph of context and a final line starting with "Decision needed".

Review a long contract or proposal before signing

Act as a contract analyst preparing a summary for an executive (you do not give legal advice).
Context: I am attaching the proposal or contract from [vendor] for [amount and term]. [Attach document]
Task: list the clauses that affect cost, timelines, penalties, automatic renewal and exit. Flag the ones that depart from the usual and the questions I should ask the vendor.
Constraints: cite the section for each finding; if something is ambiguous, say so instead of interpreting it. End by reminding me that legal must review it.
Format: table with clause, section, why it matters and suggested question.

Prepare difficult feedback

Act as a leadership coach with experience in difficult conversations.
Context: someone on my team, [role and tenure], has missed deadlines on two projects in a row: [project A] and [project B]. Until now their work was [reliable]. I do not know if there is a reason behind it.
Task: structure the conversation as opening, specific facts, listening and an improvement agreement, with an example sentence for each part.
Constraints: clear, non-accusatory tone, no assumptions about their reasons, and one open question before proposing solutions.
Format: a 20-minute script with the key sentences and two possible reactions from the person with how to respond.

Internal announcement of an organizational change

Act as a director of internal communications.
Context: I am going to tell [N people] that [two teams are merging and one of them gets a new lead]. What is already final: [dates and decisions]. What we do not know yet: [open points].
Task: write the announcement, making clear what changes, what does not change and when there will be more information.
Constraints: maximum 250 words, no euphemisms or triumphant tone, and no promises beyond what is final.
Format: email subject line, body text and three FAQs with answers.

Copilot prompts for executives

Copilot works inside Word, Excel, Outlook and Teams with the files, emails and meetings you already have access to. Check which features are active in your organization and respect each document’s permissions.

Copilot prompts for Excel, Word, Outlook and Teams

Executive summary of the monthly close (Excel and Word)

Act as a financial controller.
Context: the Excel sheet [file name] holds the close data for [month] against budget.
Task: write in Word a one-page executive summary for senior management with results versus budget, three flags and one recommendation.
Constraints: use only the figures in the sheet, state which cells or tabs each figure comes from and do not round beyond one decimal.
Format: one page, with a five-row table at the top and the text below.

Summarize a long email thread (Outlook)

Act as my executive assistant.
Context: summarize the email thread with the subject "[subject]" between [people or departments].
Task: tell me in five lines where the issue stands, what has been decided, what is still open and what is expected of me and by when.
Constraints: leave out greetings and closed details, and flag it if two people say contradictory things.
Format: three blocks titled "Decided", "Open" and "Asked of me".

Minutes and actions from a meeting (Teams)

Act as the meeting secretary.
Context: use the transcript of the Teams meeting "[meeting name]" from [date].
Task: produce the minutes with decisions made and a list of actions with owner and due date. If no owner or date was mentioned, write "unassigned".
Constraints: do not invent commitments that were not stated in the meeting and leave out side conversations.
Format: half-page minutes and an action table with task, owner and date.

Turnover patterns in a table (Excel)

Act as an HR data analyst.
Context: the table in the sheet [name] has the departures of the last six months with department, tenure and reason.
Task: calculate the turnover rate by department, identify where it is rising and propose three hypotheses about the causes.
Constraints: explain any formulas you use and mark what is data and what is interpretation; do not point at individuals.
Format: one summary table, a bar chart by department and three hypotheses in text.

Gemini prompts for executives

Gemini fits if your team works in Google Workspace, with Docs, Sheets and Gmail. Check with your administrator which Gemini features are available on your plan.

Gemini prompts for Gmail, Docs and Sheets

Prioritize this week’s inbox (Gmail)

Act as an executive assistant.
Context: review the unread emails in my inbox from [the last five days]. I lead [the operations department] and this week my priorities are [two or three priorities].
Task: group the emails into "Needs my decision", "Delegable" and "Informational", and for each of the first two groups suggest the next action.
Constraints: do not draft replies yet and flag any deadline that falls within 48 hours.
Format: three lists with sender, subject and next action in one line.

Monthly report draft in Docs (Docs)

Act as a chief of staff.
Context: in this document I have my loose notes for [month] about [department]. [Paste or reference notes]
Task: turn them into a monthly report for [my CEO] with achievements, deviations, risks and next month's priorities.
Constraints: maximum one page, short sentences, and leave any missing data in brackets instead of inventing it.
Format: four sections with headings and bullets.

Quarterly objectives tracker (Sheets)

Act as an expert in objectives and key results (OKRs).
Context: I need a sheet to track this quarter's objectives for my [department] with [N people]. The main objectives are: [objective 1], [objective 2] and [objective 3].
Task: design the sheet structure with columns, progress formulas and a view that lets me see in 30 seconds what is going wrong.
Constraints: no macros, and formulas that someone without technical experience can maintain.
Format: list of columns with their formula and an example with three rows.

More AI prompts for executives: bottlenecks, library and marketing

These prompts work the same in any tool. The first two close a diagnosis almost nobody runs: spotting bottlenecks in your team’s workflow.

Prompts to spot bottlenecks with AI

Where the work stalls

Act as a process analyst with experience in team workflows.
Context: I am sending you the status-change dates for [N] tasks from the last three months, exported from [Jira, Asana or Trello]. [Paste data]
Task: calculate how long each task spends in each status. Tell me which status accumulates the most waiting time and whether that time is concentrated in one type of task, one phase or one point in the month.
Constraints: do not tell me who is slow; tell me where the work stalls.
Format: table of average time per status and one paragraph with the conclusion.

Three process changes to reduce waiting

Act as a process improvement consultant.
Context: from the previous analysis, the status with the most waiting buildup is [status name] and it mostly affects [type of task].
Task: propose three process changes that would reduce that waiting. For each one, tell me what would need to stop happening and what signal would show, within two weeks, that the change is working.
Constraints: changes I can apply with no extra budget and without growing the team.
Format: table with change, what to stop doing and two-week success signal.

Prompts for your prompt library

Turn a prompt that works into a library entry

Act as a knowledge management lead.
Context: this is a prompt that has worked well on my [department] team: [paste prompt]. And this is an example of a good output: [paste result].
Task: turn it into an entry for our prompt library with: the situation of use written for someone brand new, the prompt with its brackets marked, the example output and the data that must never be pasted.
Constraints: plain language, no technical jargon and a maximum of 150 words excluding the prompt.
Format: a card with four sections and a final field "Reviewed on [date] by [person]".

Marketing and sales prompts

Three target audience profiles

Act as a B2B marketing strategist.
Context: we sell [description of the product or service] to [type of company or industry]. [Paste product description]
Task: define three target audience profiles with their pain points, objections and preferred channels.
Constraints: base each profile on real buying roles (who decides, who influences) and do not invent statistics.
Format: one card per profile with four fields: who they are, pain point, objection and channel.

Post ideas for your department’s LinkedIn page

Act as a LinkedIn content lead.
Context: I lead [the X department] at [company] and this quarter's content strategy is [topic or message]. The audience is [executives at mid-sized companies].
Task: propose five post ideas that maximize reach among that audience and add for each one the digital marketing angle that justifies it.
Constraints: no promotional tone, no clichés and one real data point or example in each idea.
Format: numbered list with first-line hook, core idea and a call to join the conversation.

How do you write an effective AI prompt: the five fields

A prompt is the written instruction you give an artificial intelligence tool to get a specific response. In a leadership role, that response needs to be a deliverable: a summary you can forward, a table you can put on a screen, a conversation script you can follow. These are the five fields I use in every prompt in this library.

A weak prompt turned into an effective one

Compare two versions of the same request. The weak one returns generic text you cannot take into any meeting:

Help me with tomorrow's meeting.

The effective one carries the five fields and returns a usable agenda:

Act as a project manager.
I have a check-in tomorrow with my development team and the project is a week behind schedule.
Prepare a 30-minute agenda that addresses the causes without assigning blame and ends with three concrete decisions.
Do not propose more meetings.
Format: numbered list with time allotted per item.

The second version does not require technical knowledge. It requires clarity about what you need, which is the same thing you already do when you delegate a task to someone.

Five tips for writing your own prompts

1. Start with the output format: if you know how you want to receive the answer, the rest of the prompt writes itself.
2. Paste the real data and keep brackets only for what changes each time. A prompt with generic data returns generic answers.
3. Always ask it to separate data from interpretation. It is the constraint that most improves the reliability of analysis. Close with an actionable insight (“what should I decide with this”) and validate the figures against the source.
4. Iterate with one change per message: if you do not like the result, adjust one field and compare instead of rewriting everything.
5. Save the prompts that work, with an example of a good output. That is the seed of your prompt library.

How the prompt changes in ChatGPT, Gemini, Copilot and Claude

• ChatGPT: understands full sentences and accepts files. It works well for open-ended analysis and drafts; paste the data in the message or attach the file and ask for the exact format.
• Gemini: fits Google Workspace. Ask for the action inside Docs, Sheets or Gmail and name the file or thread you mean.
• Copilot: works with the documents, emails and meetings you already have access to. Name the file or the meeting and ask for the deliverable in Word, Excel, Outlook or Teams.
• Claude: a good option with long documents and sensitive texts. Attach the whole document and ask it to cite the section behind each conclusion.

The prompting logic is the same in all four. What you learn in one tool carries over to the others, so the choice comes down to where your documents live and what guarantees you need for your data. For a wider comparison, see our guide to AI tools for executives. Each tool’s capabilities change frequently: check current features before deciding.

How to build a prompt library for your leadership team

A prompt library is a shared document holding the instructions that have already worked, saved in a way someone else can use without asking you first. It is what turns an individual skill into a team capability. It is also the thing most likely to get abandoned: most die within three weeks by turning into a dumping ground of text with no context. What separates the ones that survive is that each entry keeps four things, not just the prompt text:

1. The specific situation it is used for, written the way someone brand new would describe it, which also helps onboarding new managers.
2. The full prompt, with the brackets that need to be swapped clearly marked.
3. A real example of good output, so the next person knows what they are aiming for.
4. The date of the last review and who is responsible for that entry.

Add two governance rules. One person per department reviews their section once a quarter and deletes what nobody uses anymore. And the header of the document carries the list of data that never gets pasted in, so the rule lives where the decision gets made instead of in a policy nobody opens. When several teams share a library, instructions stop being rewritten from scratch and improve version over version.

Limits of generative AI and confidential data

AI does not replace the judgment of the person in charge, and there are three areas where human oversight is not optional.

1. Personal, financial or strategic data in tools whose data-handling policy you do not know. Check whether you are on a business plan or a free one before you paste anything in.
2. Decisions that affect people. A model can compare scenarios; it does not take responsibility for a promotion, a termination or a performance review.
3. Model bias. It reproduces the patterns in its training data, and in hiring or evaluation that has measurable consequences.

Frequently asked questions about AI prompts for executives

These are the questions people ask most about AI prompts and ChatGPT prompts for executives, with the answer first.

What is a prompt, and what is prompting?

A prompt is the written instruction you give a tool like ChatGPT, Copilot, Gemini or Claude to get a useful response. Prompting is the craft of writing those instructions well. For leadership use, a good prompt includes a role, context, task, constraints and an output format.

How do you write an effective AI prompt?

Define five things: the role the model should take, the context of your situation, the concrete task, the constraints and the format of the answer. Avoid instructions like “help me with this.” Structure matters more than length.

What is the best AI for executive prompts?

There is no single best. ChatGPT is good for open-ended analysis, Claude for long documents and sensitive texts, Copilot if you work in Word, Excel, Outlook and Teams, and Gemini if your company uses Google Workspace. The prompt logic is the same in all of them.

Which ChatGPT prompts work for executives?

The ones that target high-volume, repetitive tasks: summarizing reports, comparing scenarios, preparing check-ins and spotting bottlenecks in a workflow. You have six ready to copy in the ChatGPT section.

How do I use Copilot as an executive?

Copilot works inside Word, Excel, Outlook and Teams with the files, emails and meetings you already have access to. Name the file or the meeting in the prompt and ask for a concrete deliverable, such as a thread summary or minutes with actions.

Does Gemini work for executive prompts?

Yes, especially if your team works in Google Workspace. It helps prioritize your Gmail inbox, draft reports in Docs and build tracking sheets in Sheets. Available features depend on your plan.

Can I use Claude for executive prompts?

Yes. Claude works well with long documents, such as contracts or technical reports, and with sensitive texts, such as an internal announcement or difficult feedback. Attach the whole document and ask it to cite the section behind each conclusion.

How do you manage and track a team’s prompts?

With a shared prompt library: one entry per recurring task with the situation of use, the full prompt, an example of a good output and the review date with its owner. One person per department reviews it each quarter and the header lists the data that never gets entered.

Are AI tools safe for confidential company data?

It depends on the tool and how it is configured. Some free versions may use what you enter to improve their models, while business plans usually come with different contractual guarantees. Check the privacy policy and keep personal, financial and strategic data out until you confirm how it is handled.

Your next step with AI prompts for executives

Pick one of the prompts in the library, change what is in brackets and run your first test this week on a real task. Then change a single field, such as the format or the constraints, and compare. That teaches you what each field controls much faster than reading theory.

If you want to apply this with a method, Founderz’s AI & Innovation Program, developed in partnership with Microsoft, is built for working professionals: fully online and applied to real work. Founderz is an online business school specialized in applied AI.