Book A Free Intro Call
The Oxford AI School

Home/AI Learning Hub/What Are The Best Prompts For.../Managers

What Are the Best AI Prompts for Managers? 10 to Copy and Use

Ten prompts for line managers and team leaders: 1:1 prep, feedback, delegation, fair performance reviews, meetings, workload, difficult conversations and onboarding a new starter.

By Harry LangPublished 9 November 202611 minute read

The Line Manager's Loom > best prompt for... less admin more people The Line Manager's Loom Fig. 4
In this article
  1. What are the best AI prompts for managers?
  2. How can managers use ChatGPT for 1:1s and feedback?
  3. Which AI prompts help managers delegate and plan workload?
  4. How should managers use AI for performance reviews fairly?
  5. What are good Copilot prompts for meetings and onboarding?
  6. How do managers build a team AI habit?
  7. Is a prompt list enough to make managers good with AI?
  8. AI prompts for managers: questions people ask

The short answer

The best AI prompts for managers handle the admin around people: preparing 1:1s, drafting specific feedback, writing delegation briefs, structuring fair performance reviews, planning workload, setting agendas, capturing actions and onboarding new starters. Give ChatGPT, Claude, Copilot or Gemini real context and a clear format, keep personal data out of unapproved tools, and make the judgements yourself.

The best AI prompts for managers take the admin out of management so you have more time for the people part. Used well, they help you prepare a 1:1 in five minutes, give feedback that lands and write a delegation brief nobody has to ask about twice.

Line managers are usually the busiest people in a business and the least trained. ChatGPT, Claude, Microsoft Copilot and Gemini won't fix that on their own, but the ten prompts below cover the jobs I see managers wrestle with most. They are written for UK team leaders, and every one tells the AI who it is, what you're giving it, what you want back and what to avoid.

What are the best AI prompts for managers?

The best AI prompts for managers cover 1:1 prep, feedback, delegation, performance reviews, meetings, workload planning, difficult conversations and onboarding. They work because they give the AI real context about your team and ask for a specific, usable output. Here is the set.

Ten AI prompts for managers at a glance
PromptWhen to use itWhat you get
1. 1:1 prepThe night before a 1:1Agenda, questions, follow-ups
2. Feedback that landsAfter a moment worth namingSpecific, balanced feedback
3. Difficult conversation rehearsalBefore a hard chatOpening lines and likely reactions
4. Delegation briefHanding over a taskA clear one-page brief
5. Workload plannerMonday morning or a crunch weekPriorities and trade-offs
6. Performance review draftReview seasonA structured, evidence-based draft
7. Bias checkBefore you submit a reviewLanguage and fairness flags
8. Meeting agendaBefore a team meetingA timed agenda with outcomes
9. Actions from notesAfter the meetingOwners, dates and a follow-up email
10. New starter planBefore someone joinsA first-month onboarding plan

How can managers use ChatGPT for 1:1s and feedback?

Use ChatGPT to structure your thinking before a 1:1 or a feedback conversation, then have the conversation yourself. The AI is good at turning scattered notes into a plan and at spotting when your feedback is vague. It cannot know your colleague, so you supply the facts.

Prompt 1: The 1:1 prep

Prompt 1: 1:1 prep
You are an experienced people manager helping me prepare for a 30-minute 1:1 with a member of my team.

Context: their role is [role], they have been in post [time], and their current priorities are [list]. My notes since our last 1:1: [paste notes, no personal or health details].

Give me:
1. A short agenda with timings, starting with their items.
2. Five open questions that help them talk about progress, blockers and wellbeing.
3. One thing I should recognise them for.
4. Follow-ups from last time I must not forget.

Keep it to one screen.

Why this prompt works

Starting with their items and open questions keeps the 1:1 about them, not your status update.

Prompt 2: Feedback that lands

Prompt 2: Feedback that lands
Act as a management coach. I want to give feedback to a colleague and I want it to be specific, fair and useful.

What happened: [describe the situation and the behaviour you saw, facts only]. The impact: [effect on the team, customer or work]. What I would like next time: [your request].

Write the feedback using the situation, behaviour, impact structure, in no more than 120 words, in a warm but direct British tone. Then give me two questions to ask so it becomes a conversation, not a verdict. Point out anything in my description that sounds like opinion rather than observed behaviour.

Why this prompt works

The last line catches the judgements that creep into feedback and make people defensive.

Prompt 3: Difficult conversation rehearsal

Prompt 3: Difficult conversation rehearsal
I need to have a difficult conversation with a team member about [issue, for example missed deadlines]. Here is the background: [facts, no names].

First, help me write a calm opening of two or three sentences. Then play the team member and respond realistically, including some pushback. After each of my replies, step out of role and tell me briefly what worked and what could escalate things. Stop after five exchanges and summarise what I should say and avoid.

Remind me where I should involve HR or follow our formal policy.

Why this prompt works

Rehearsing out loud with a sparring partner is the closest thing to practice most managers ever get.

Which AI prompts help managers delegate and plan workload?

AI prompts help managers delegate by turning a task in your head into a written brief, and plan workload by forcing trade-offs onto the page. Both save the time you would otherwise spend answering questions later.

Prompt 4: The delegation brief

Prompt 4: Delegation brief
You are an operations manager who writes excellent briefs. Turn my rough notes into a one-page delegation brief for [role].

My notes: [paste what the task is, why it matters, deadline, budget, who else is involved].

Use these headings: Outcome we need, Why it matters, What good looks like, Deadline and checkpoints, Decisions you can make without me, Decisions to check with me, Resources and people. Write in plain English. Then list three questions the person is likely to ask that my notes don't answer.

Why this prompt works

The "decisions you can make" heading is what turns a handover into real delegation.

Prompt 5: The workload planner

Prompt 5: Workload planner
Act as a pragmatic team lead. Here is my team's work for the next two weeks: [paste tasks, owners, deadlines and rough effort] and our capacity: [who is in, holidays, part-time hours].

Show me where we are overloaded, which tasks are at risk, and three realistic options (for example delay, drop, reassign or get help), with the trade-off of each. Put the overview in a table by person. Do not assume people can work extra hours.

Why this prompt works

The final constraint stops the AI solving every problem with overtime, which is what a tired manager does too.

How should managers use AI for performance reviews fairly?

Use AI to organise evidence and tighten your writing, never to decide the rating. A performance review affects someone's pay, progression and confidence, so the judgement must be yours and based on facts. AI can also introduce or amplify bias, so check its language as carefully as your own.

Prompt 6: Performance review draft

Prompt 6: Performance review draft
You are helping a line manager draft a fair, evidence-based performance review.

The objectives for the period: [paste]. Evidence I have collected: [examples, results, feedback, with names removed]. My overall view: [your own assessment].

Draft the review under each objective: what was achieved, the evidence, and one development point. Use only the evidence provided and do not invent examples. Use neutral, specific language and avoid words about personality. Where evidence is thin, write [NEED EVIDENCE] instead of filling the gap.

Why this prompt works

You set the judgement and the AI is boxed in to your evidence, so it can't write a confident review out of nothing.

Prompt 7: The bias check

Prompt 7: Bias check
Review this draft performance review for fairness before I submit it: [paste draft, anonymised].

Flag:
- vague or personality-based words (for example "abrasive", "lacks presence")
- language that could read differently depending on gender, age, ethnicity, disability or other protected characteristics
- claims without evidence
- recent events that may be outweighing the whole period

For each flag, quote the phrase, explain the risk in one line and suggest a fairer, evidence-based alternative. Do not change my overall assessment.

Why this prompt works

Naming the common traps, such as recency and personality words, gives the AI something concrete to look for.

What are good Copilot prompts for meetings and onboarding?

Good Copilot prompts for meetings set a purpose and outcomes before the meeting and turn notes into owned actions afterwards. If your team works in Microsoft 365, Copilot can pull from your calendar, Teams chats and documents; the same prompts work in ChatGPT, Claude or Gemini if you paste the material in.

Prompt 8: A meeting agenda that earns its slot

Prompt 8: Meeting agenda
You are a facilitator who hates wasted meetings. Help me plan a [length] team meeting.

Purpose: [why we are meeting]. Attendees: [roles]. Topics people have raised: [list].

Create an agenda with: the outcome we need from each item (decision, input or update), a time for each, who leads it, and anything people should read beforehand. Move pure updates to a written note instead. Finish with a one-line purpose statement I can put in the invite.

Why this prompt works

Asking for the outcome of each item exposes the agenda items that should have been an email.

Prompt 9: Actions from meeting notes

Prompt 9: Actions from notes
Here are the notes or transcript from today's meeting: [paste or reference the meeting in Copilot].

Extract:
1. Decisions made.
2. Actions, each with one owner and a due date (write [NO OWNER] or [NO DATE] if missing, don't guess).
3. Open questions still to resolve.

Then write a short follow-up email to attendees in a friendly, professional tone, with actions in a table. Keep the email under 200 words.

Why this prompt works

Refusing to guess owners exposes the actions that would otherwise quietly belong to nobody.

Prompt 10: A new starter's first month

Prompt 10: New starter plan
Act as an experienced manager who onboards people well. Someone is joining my team as [role] on [date]. The team does [summary]. The tools they'll use: [list]. Key people they need to meet: [roles].

Create a four-week onboarding plan: week by week, with what they should learn, who they should meet, one small real piece of work each week, and a check-in question for the end of each week. Include a first-day schedule. Keep it practical for a UK business of around [number] people.

Why this prompt works

One small piece of real work each week builds confidence faster than a month of reading.

How do managers build a team AI habit?

Managers build a team AI habit by picking a few repeat tasks, agreeing shared prompts and safe-use rules, and reviewing what works each month. A good prompt saved in one person's chat history helps one person. A good prompt shared with the team helps everyone.

  1. Pick three repeat jobs.Choose tasks your team does every week, such as meeting notes or customer replies.
  2. Write one shared prompt for each.Use the structure on this page and store them somewhere everyone can find.
  3. Agree the rules.Which tools are approved, what data stays out, and when a human checks the output. Your AI policy should answer this.
  4. Practise on real work.Spend 30 minutes in a team meeting running a prompt on a live task and comparing results.
  5. Review monthly.Keep what saves time, fix what doesn't, and add the next job.

Here is the difference a little structure makes:

The lazy version
Hey, can you write me an email to the team about the new rota changes that are coming in next month, making sure it is clear and friendly and covers everything they need to know about it so there are no complaints? [paste rota details]

Why it fails: It asks for "everything" without structure, so you get a long email nobody reads.

The better version
You are a team leader at a UK [sector] business. Write a short email to my team of [number] about rota changes starting [date]: [paste details]. Lead with what changes and why, then what each person needs to do, then who to ask. Friendly, plain English, under 180 words. Add a two-line summary for Teams.

Why this prompt works

It sets the order, the length and the reader, so the email does one job well.

For more on structure, read how to write prompts for AI, and if you want a wider bank of everyday prompts, ChatGPT prompts for work is a good companion.

Is a prompt list enough to make managers good with AI?

No. A list gets a manager started, but the gains come when a whole team shares habits, follows the same policy and has practised on its own work. Managers are also the people who decide whether AI use spreads or stalls in their team, so they need confidence, not just a cheat sheet.

That is what my training is for. Basic AI Training gives a team the foundations, and the Microsoft Copilot Deep Dive suits teams living in Microsoft 365. If you're choosing, my guide to AI courses for managers sets out the options, and the AI prompts for team AI training page shows how to run a first session. Directors overseeing all this should see the AI prompts for directors.

AI prompts for managers: questions people ask

What are the best AI prompts for managers?

The most useful cover 1:1 prep, feedback, delegation briefs, workload planning, performance reviews, meeting agendas, actions from notes and onboarding plans. A good prompt gives the AI a role, your context, a clear task, an output format and limits. The ten on this page are a solid starting set.

Can managers use ChatGPT to write performance reviews?

You can use it to organise your evidence and tighten wording, but the assessment must be yours and based on facts. Anonymise what you paste, tell the AI not to invent examples, and run a bias check on the draft before you submit it.

Is it safe to put 1:1 notes into AI?

Only in a tool your organisation has approved, and ideally with names and sensitive details removed. 1:1 notes often contain personal data covered by UK GDPR. See how to protect business data when using AI.

What are good Copilot prompts for managers?

Copilot is strongest where it can see your Microsoft 365 data: summarising a Teams meeting, extracting actions with owners, drafting a follow-up email or preparing for a 1:1 from recent emails and chats. Prompts 8 and 9 on this page work well in Copilot.

Can AI help with difficult conversations at work?

Yes, as a rehearsal partner. Ask it to play the other person, push back realistically and give you feedback on your wording. It does not replace your HR policy or advice on formal processes.

Should managers have AI training?

Yes, because managers decide whether AI use spreads safely in their teams. Short, practical training on real work builds the habits a prompt list can't. My guide to AI courses for managers compares the options.

Keep reading

The Oxford AI School crest

Give your managers AI habits that stick

I train management teams on their own real work, in person or online, so they leave with shared prompts, safe-use rules and the confidence to bring their teams along.

Harry Lang
Harry Lang founded The Oxford AI School in Witney, Oxfordshire. He trains UK businesses and teams to use AI well, writes the Marketing Unfiltered newsletter, and is the author of Learn To Fly With AI.