Lesson 5 of 6·7 min read

Prompts for Analysis and Reporting 📊

AI can not only process data — it can interpret, contextualize, and explain it. This makes it the ideal analysis partner, generating actionable insights from raw data.

🎯 What You'll Learn

  • Prompt templates for trend, competitive, and financial analysis
  • Report templates for different audiences
  • Best practices for analysis prompts
  • Data security in AI-powered analysis

Trend Analysis 📈

Role: You are a Senior Business Analyst.

Here are our revenue figures for the last 12 months:
[INSERT DATA — e.g., as a table or CSV]

Analyze:
1. Overall trend (rising/falling/stagnating) with growth rate
2. Seasonal patterns and their possible causes
3. Notable outliers — what might have caused them?
4. Forecast for the next 3 months (conservative/realistic)

Format: Executive summary (max 200 words), then detail table.
Audience: Executive management.

Competitive Analysis 🔍

Compare the following three competitors:
- [COMPETITOR 1]: [Info/URL]
- [COMPETITOR 2]: [Info/URL]
- [COMPETITOR 3]: [Info/URL]

Criteria: Price, core features, target audience, market
position, strengths, weaknesses.

Deliver:
1. Comparison table (6 criteria × 3 competitors)
2. SWOT analysis for US compared to them
3. Top 3 differentiation opportunities

Context about us: [BRIEF COMPANY DESCRIPTION]

Financial Analysis 💰

Role: You are a Financial Controller.

Analyze the following quarterly figures:
[INSERT DATA]

Create:
1. Budget vs. actual comparison with variance analysis
2. Identify top 3 cost drivers
3. Cash flow trends and liquidity forecast
4. Action recommendations (prioritized by impact)

Format: Management dashboard style with numbers, arrows
(↑↓→) and brief explanations. Max 1 page.

💡 Tip: Always include the comparison period for the AI. "Compare Q1 2026 with Q1 2025 and Q4 2025" delivers significantly better results than "analyze our numbers."

Report Templates for Different Audiences 👥

Report TypePrompt StrategyAudience
Executive Summary"Summarize in 5 bullet points, max 100 words"Executive management
Detail Report"Analyze thoroughly with data basis and methodology"Department teams
Status Report"Create in traffic light format: Green/Yellow/Red with explanation"Project management
Customer Analysis"Segment by behavior, value, and churn risk"Sales/Marketing
Board Report"Strategic highlights with trends and risks, max 1 page"Board of directors

🔑 Remember: Always adjust the audience in the prompt. A report for executives looks completely different from one for the operational team.

Best Practices for Analysis Prompts 🎯

  1. Always include data — AI can't guess. Copy tables directly into the prompt or use file upload features
  2. Provide context — Industry, company size, time period, previous results
  3. Specify format — Table, bullet points, executive summary, dashboard style
  4. Define audience — "For the board" produces different results than "for the operational team"
  5. Set boundaries — "Base your analysis exclusively on the provided data, no assumptions"
  6. Build in queries — "If you're missing important data, say which and why"

⚠️ Caution: Before uploading, check whether sensitive data needs to be anonymized or aggregated. Use enterprise versions with data privacy guarantees (e.g., ChatGPT Enterprise, Claude for Business). Never enter personal data into public AI tools.

Dashboard Generation 📊

Modern AI tools like Claude Opus 4.6 and GPT-5 can even generate interactive dashboards:

Create an HTML dashboard with the following KPIs:
[KPI LIST]

Use Chart.js for visualizations:
- Revenue trend as line chart
- Cost distribution as pie chart
- Top 5 products as bar chart

Data: [INSERT DATA]
Design: Minimalist, dark mode, responsive.

🏢 Real-world example: A sales director has Claude create a pipeline analysis every Monday. Input: CRM export as CSV. Output: Prioritized deal list with win probabilities and recommended next steps. Time saved: 2 hours per week.

📋 Summary

  • Analysis prompts need: Data + Context + Format + Audience
  • Use specific templates for different analysis types
  • Adjust the output to the audience (Board vs. Team)
  • Respect data privacy: anonymize sensitive data, use enterprise versions
  • AI can also generate dashboards and visualizations

🎯 Exercise: Export a dataset from your daily work (e.g., revenue, customer data, project hours) and create a management report with AI. Compare the time investment.

Next lesson: Prompt Chaining and Structured Output — connecting multiple prompts into automated workflows.

Previous lessonNext lesson