The difference between a mediocre AI experience and a transformative one is almost always the prompt. A 2025 study by Harvard Business School found that professionals who used structured prompting techniques produced outputs rated 43% higher in quality than those who typed simple, conversational requests. The tool is the same. The input makes the difference.
This guide provides 25 tested, copy-paste-ready prompts for the most common business functions. Each prompt follows principles that consistently produce better results: clear role assignment, specific context, defined output format, and explicit constraints. These work with ChatGPT, Claude, Gemini, and most other AI tools — the principles are universal.
À retenir
- Structured prompts produce outputs rated 43% higher in quality than ad-hoc requests
- The five elements of an effective prompt: role, context, task, format, constraints
- 25 prompts organised by function: HR, marketing, finance, legal, and management
- Every prompt is copy-paste ready — customise the bracketed sections for your context
The anatomy of a good prompt
Before the prompts, a 30-second primer on why they work. Every effective business prompt includes:
- Role: Tell the AI who it is. “You are a senior HR business partner with 15 years of experience” produces vastly different output from “Help me with HR.”
- Context: Provide relevant background. Industry, company size, audience, constraints.
- Task: Be specific about what you want. “Analyse” is vague. “Compare these two candidates against these five criteria and recommend one with reasoning” is specific.
- Format: Define the output structure. Table, bullet points, email, report section, numbered list.
- Constraints: Set boundaries. Word count, tone, what to avoid, what to include.
For a deeper dive into these principles, see our prompt engineering guide.
43%
higher quality ratings for outputs produced with structured prompts vs. unstructured requests
Source : Harvard Business School, AI Productivity Study 2025
HR prompts (1–5)
These prompts are designed for recruitment, people management, and employee development. If your organisation is considering using AI in hiring or performance evaluation, note that these may fall under the EU AI Act’s high-risk category — always check your obligations.
1. Write a job description
You are a senior HR specialist writing a job description for a [job title] at a
[industry] company with [number] employees. The role reports to [manager title]
and is based in [location/remote].
Write a job description that includes: role summary (3 sentences), 5-7 key
responsibilities, required qualifications (5 max), preferred qualifications
(3 max), and a brief section on company culture.
Tone: professional but approachable. Avoid clichés like "rock star" or "ninja."
Use gender-neutral language throughout. Length: 400-500 words.2. Screen a CV against job requirements
You are a recruitment specialist. I will provide a job description and a CV.
Analyse the CV against the job requirements and produce:
1. A match score (0-100%) with justification
2. A table: requirement vs. evidence found in CV vs. assessment (Strong/Partial/Missing)
3. Three specific interview questions to explore gaps
4. A recommendation: proceed to interview / hold / reject — with reasoning
Job description: [paste]
CV: [paste]3. Draft a performance review
You are an experienced people manager. Draft a balanced performance review for
[employee name/role] covering the period [dates]. Use this structure:
- Overall assessment (2-3 sentences)
- Key achievements (3-4 bullet points with specific outcomes)
- Areas for development (2-3 bullet points — constructive, actionable)
- Goals for next period (3 SMART goals)
Context: [add context about the employee's role, team, key projects].
Tone: honest, supportive, specific. Avoid vague praise like "great team player"
— always link to specific actions or outcomes.4. Create an onboarding plan
Design a 30-day onboarding plan for a new [job title] joining a [industry]
company. The plan should cover:
- Week 1: orientation, systems access, key meetings
- Week 2: role-specific training, first tasks, buddy assignment
- Week 3-4: increasing independence, first deliverables, feedback checkpoints
Format as a table with columns: Day/Week | Activity | Owner | Success Criteria.
Include a 30-day review checklist at the end.5. Handle a difficult conversation
You are an HR coach. I need to have a conversation with an employee about
[situation: e.g., consistent lateness, performance gap, interpersonal conflict].
Help me prepare:
1. Opening statement (direct but empathetic, 2-3 sentences)
2. Key points to cover (bullet format)
3. Anticipated responses and how to handle them
4. Desired outcomes and next steps
5. Phrases to avoid and better alternatives
Context: [relationship history, previous conversations, severity].Marketing prompts (6–10)
6. Write a LinkedIn post
You are a B2B content strategist. Write a LinkedIn post for [company/person]
about [topic]. Target audience: [describe]. Goal: [awareness/engagement/traffic].
Structure: hook (first line must stop scrolling), 3-4 short paragraphs,
a clear call to action. Use line breaks for readability.
Tone: [conversational/authoritative/provocative]. Length: 150-200 words.
Do not use hashtags in the body — add 3-5 relevant hashtags at the end only.7. Create an email campaign sequence
You are a senior email marketing specialist. Create a 5-email nurture sequence
for [product/service] targeting [audience]. Goal: move prospects from awareness
to booking a demo.
For each email provide: subject line (under 50 characters), preview text,
body copy (150-200 words), and CTA. The sequence should progress:
Email 1: problem awareness | Email 2: solution introduction | Email 3: social
proof | Email 4: objection handling | Email 5: urgency and final CTA.
Tone: [describe]. Avoid: spam trigger words, excessive exclamation marks,
false urgency.8. Analyse competitor positioning
Analyse the positioning of these competitors: [list 3-5 competitors with URLs].
For each, identify:
1. Core value proposition (1 sentence)
2. Target audience
3. Key messaging themes (3-5)
4. Pricing model
5. Differentiators vs. [your company]
Then summarise: gaps in the market we could exploit, messaging angles competitors
are not using, and positioning risks to watch.
Format: summary table followed by strategic analysis (300 words max).9. Repurpose content across channels
I have a [blog post/report/presentation] about [topic]. Repurpose it into:
1. A LinkedIn post (150 words)
2. A Twitter/X thread (5 tweets)
3. An email newsletter snippet (100 words)
4. Three Instagram carousel slide texts (30 words each)
5. A 60-second video script
Maintain the core message but adapt tone and format for each channel.
Source content: [paste]10. Write website copy for a product page
You are a conversion-focused copywriter. Write product page copy for [product]
targeting [audience]. Include:
- Headline (under 10 words, benefit-focused)
- Subheadline (1 sentence, expand on the headline)
- Three feature blocks: feature name, 2-sentence description, specific benefit
- Social proof section placeholder
- CTA section with primary and secondary CTA text
Tone: clear, confident, no jargon. Follow the "so what?" test — every feature
must link to a tangible benefit.2-3x
more value extracted from AI tools by trained employees compared to untrained colleagues performing identical tasks
Source : MIT Sloan / Stanford, AI Productivity Research 2024
Finance prompts (11–15)
AI is increasingly used for financial analysis, but output verification is critical. These prompts produce useful first drafts — never treat AI-generated financial data as final without human review. For guidance on which AI tools work best for financial tasks, see our comparison guide.
11. Summarise a financial report
Summarise this financial report for a non-finance audience. Extract:
1. Top 3 key findings (1 sentence each)
2. Revenue and profit trends (with % changes)
3. Red flags or areas of concern
4. Comparison to [previous period/industry benchmarks]
Format: executive summary (200 words max) followed by a data table.
Report: [paste]12. Create a budget variance analysis
Analyse this budget vs. actual data and produce:
1. A variance table (category, budget, actual, variance £, variance %)
2. Top 5 variances ranked by materiality
3. Likely causes for each significant variance
4. Recommended actions
Data: [paste budget and actuals]13. Draft a business case
Draft a business case for [investment/project]. Include: executive summary,
problem statement, proposed solution, cost-benefit analysis (3-year horizon),
risks and mitigations, implementation timeline, recommendation.
Context: company size [X], industry [Y], budget range [Z].
Length: 800-1000 words. Format: professional report structure with headers.14. Explain a financial concept to non-experts
Explain [financial concept: e.g., working capital, EBITDA, cash flow forecasting]
to a non-finance audience. Use: 1 clear definition (2 sentences), 1 real-world
analogy, 1 worked example with simple numbers, and 1 paragraph on why it matters
for business decisions. Avoid jargon. Length: 200 words.15. Review an invoice for anomalies
Review this invoice against our purchase order and flag any discrepancies:
quantity differences, pricing variances, unexpected charges, missing information,
or terms that differ from our standard agreements.
Invoice: [paste] | Purchase order: [paste]
Format: table with columns: Line Item | Expected | Invoiced | Discrepancy | ActionLegal prompts (16–20)
16. Summarise a contract
Summarise this contract for a non-legal audience. Extract:
1. Parties and effective date
2. Key obligations (each party, bullet points)
3. Payment terms and amounts
4. Termination clauses and notice periods
5. Liability and indemnity provisions
6. Unusual or potentially problematic clauses (flag these)
Length: 300 words max. Use plain language throughout.
Contract: [paste]17. Compare two contract versions
Compare these two versions of [contract type] and produce a redline summary:
1. Table of changes: clause, old text, new text, significance (low/medium/high)
2. Assessment of who benefits from each change
3. Recommended negotiation points
Version A: [paste] | Version B: [paste]18. Draft a simple policy
Draft an [AI usage/data protection/remote working] policy for a [size] company
in [industry]. Include: purpose, scope, definitions, rules and guidelines,
responsibilities, compliance monitoring, breach consequences, review schedule.
Tone: clear and direct. Length: 600-800 words.
Ensure compliance with [relevant regulations: e.g., GDPR, EU AI Act].19. Assess regulatory impact
Analyse how [specific regulation: e.g., EU AI Act Article 4] affects a [size]
company in [industry]. Cover: which obligations apply, compliance deadline,
practical steps required, estimated effort/cost, penalties for non-compliance.
Focus on practical impact, not legal theory. Length: 400 words.20. Prepare meeting minutes
Convert these raw meeting notes into formal minutes. Include: date, attendees,
agenda items discussed, key decisions (numbered), action items (owner + deadline),
and next meeting date.
Notes: [paste]. Format: professional minute template.AI tools can help with legal tasks but must never replace qualified legal advice. Always have a lawyer review AI-generated contract analysis, policy drafts, and regulatory assessments. AI models hallucinate legal provisions that do not exist. See our trustworthy AI guide for more on managing AI limitations.
Management prompts (21–25)
Leadership communication and strategic analysis are areas where AI adds genuine value — provided you bring the context and judgement. For organisations undergoing AI transformation, these prompts support the change management process.
21. Prepare a strategy briefing
Create a strategy briefing on [topic] for a senior leadership team. Include:
current situation (3-4 key facts with data), market trends, competitive
landscape, three strategic options (with pros/cons), and a recommended path.
Context: [company size, industry, current position]. Length: 500 words.22. Facilitate a decision
I need to decide between [Option A] and [Option B] for [context]. Act as a
strategic advisor. Create: a comparison matrix (criteria: cost, time, risk,
strategic fit, team impact), a devil's advocate argument for each option,
and a recommended decision with 3 supporting reasons.23. Write a change communication
Draft an internal communication announcing [change: e.g., new AI tools rollout,
restructuring, policy change] to all employees. Include: what is changing and
why, what it means for employees, timeline, support available, and FAQ section
(5 anticipated questions with answers).
Tone: transparent, empathetic, confident. Length: 400 words.24. Create a project status report
Convert these project notes into a structured status report:
- Project: [name] | Period: [dates] | PM: [name]
- Status: [Green/Amber/Red] with justification
- Achievements this period (bullet points)
- Risks and issues (table: description, impact, probability, mitigation)
- Next period priorities
- Budget status: planned vs. actual
Notes: [paste]25. Analyse employee survey results
Analyse these employee survey results and produce:
1. Executive summary (key themes in 3-4 sentences)
2. Top 3 strengths (with supporting data)
3. Top 3 concerns (with supporting data)
4. Comparison to [previous survey/benchmarks] if provided
5. Three specific, actionable recommendations
Data: [paste]. Prioritise insights that are actionable over those that are merely interesting.Save your best prompts as templates. Build a shared prompt library for your team so everyone benefits from what works. Over time, this becomes one of your most valuable AI assets. For a structured approach to building AI competency across your organisation, see our AI competency framework guide.
From prompts to competency
Good prompts are the starting point. Real AI competency means understanding when to use AI, when not to, how to verify outputs, and how to handle data responsibly. Brain provides structured training that takes your team from copying prompts to genuinely understanding how to work with AI — covering prompt engineering, output verification, AI governance, and EU AI Act compliance. See our plans to get started.
Related articles
AI for Meetings: Save 5+ Hours Per Week (2026)
Make every meeting count with AI transcription, summaries, action items, scheduling, and analytics. Includes data privacy and security guidance.
ChatGPT for Business: Enterprise Adoption Guide 2026
Deploy ChatGPT across your organisation successfully. Tier comparison, department use cases, data privacy and team training covered.
Generative AI Business Roadmap: Strategy Guide 2026
Plan your generative AI rollout with use cases by department, vendor comparisons, risk guardrails, and a cost-benefit framework.