Analyze Spreadsheets with AI: the Copy-Paste Prompt
You open the Excel file with the quarterly numbers. 847 rows, 12 columns. Somewhere in this sea of numbers lies the answer to your boss's question: 'Why did the numbers in the South region collapse?' You start sorting, filtering, building pivot tables -- and after an hour you have three charts but no real explanation.
The core problem: We have plenty of data -- insights are what's missing.
Anyone who works with spreadsheets knows this feeling. The raw data is there: revenue figures, survey results, website analytics, customer feedback, project hours. But from data point to insight is a long road. Manually spotting patterns, finding correlations, identifying outliers -- it takes hours, and you still miss things.
The solution: AI as your data analyst.
Claude, ChatGPT, and Gemini can analyze tabular data in seconds. Not superficially ('the numbers are rising'), but with real insights: trends over time, unusual outliers, relationships between columns, and -- most importantly -- concrete, actionable recommendations.
How does it work in practice?
You copy your data directly from Excel, Google Sheets, or a CSV file and paste it together with the prompt into your AI tool. The AI automatically recognizes column structure, data types, and relationships.
How to prepare your data:
1. From Excel/Google Sheets: Select the range (including column headers), copy it (Ctrl+C), and paste it directly into the chat. The table structure is preserved -- columns are separated by tabs.
2. From a CSV file: Open the file in a text editor, copy the contents, and paste.
3. Large datasets: Most AI tools handle 500-1000 rows in chat without issues. For larger datasets: Upload the file (Claude and ChatGPT support CSV/Excel upload) or work with a representative sample.
Which data works especially well?
- Revenue and sales figures: Trends, seasonal patterns, product performance
- Survey results: Satisfaction scores, free-text categorization, correlations
- Website analytics: Traffic sources, conversion rates, bounce rates
- HR data: Absences, turnover, team sizes (mind privacy!)
- Project data: Time spent, budget variances, milestone achievement
- Customer feedback: Ratings, complaints, support tickets
Privacy note:
Before pasting company data into an AI tool:
- Personal data: Names, email addresses, salaries do not belong in external AI tools. Anonymize or pseudonymize before pasting (Person 1, Person 2, or Department A, B, C).
- Business-critical numbers: Check whether your company has policies for using external AI tools. Many companies allow aggregated data but not raw data with customer names.
- Use enterprise AI: Microsoft Copilot, Google Duet AI, or Claude for Business process data within the corporate environment. Ask your IT department.
- Aggregate instead of raw data: Often summaries (totals per region, averages per month) are sufficient instead of individual records.
Three scenarios showing what's possible:
Scenario 1 -- Preparing a quarterly review:
You have revenue data from 4 regions over 12 months. Manually, you would need 2 hours for a solid analysis. With the prompt, you get in 2 minutes: which region is growing, which is stagnating, where seasonal patterns exist, and three concrete recommendations for the next quarter.
Scenario 2 -- Evaluating an employee survey:
The annual survey has 200 responses with 15 questions (scale 1-5 plus free text). Instead of spending weeks building pivot tables and reading free-text responses, you paste the results into the prompt. You get: the 3 strongest and 3 weakest areas, surprising correlations ('teams with high leadership satisfaction also have lower absence rates'), and prioritized action items.
Scenario 3 -- Understanding website performance:
Your Google Analytics export shows 6 months of traffic data. You see that traffic has increased but the conversion rate has dropped. Why? The prompt identifies: the traffic increase comes mainly from social media with low purchase intent, while organic traffic (with high conversion) is stagnating. Recommendation: invest in SEO instead of social media campaigns.
Pro tips for better analyses:
- Provide context: Write before the data what it is about: 'This is monthly revenue data from a mid-size SaaS company. There was a 15% price increase in March.' The more context, the more relevant the insights.
- Ask specific questions: Add to the prompt: 'Also answer this question: Is there a correlation between team size and project duration?' Targeted questions produce targeted answers.
- Request visualizations: After the analysis, say: 'Create the Python code for the 3 most important charts I should show my boss.' Or: 'Which Excel charts would you create from this data?'
- Work iteratively: Start with the overall analysis, then zoom in: 'Now analyze only the South region in detail' or 'Compare the top 3 products with each other.'
- Generate formulas: 'Give me the Excel formulas to replicate this analysis myself.' Perfect for learning and recurring analyses.
- Draft reports: 'Summarize the 3 most important findings in one paragraph I can paste directly into my quarterly report.'
You are an experienced data analyst. Your task is to thoroughly analyze the following tabular data and derive clear, actionable insights. **Context (optional but recommended):** - Data source: [e.g., revenue data Q1-Q2 / employee survey 2026 / website analytics last 6 months] - Industry/area: [e.g., SaaS company / retail / internal HR team] - Special circumstances: [e.g., price increase in March / new product since April / team restructuring] - Specific question: [e.g., Why is revenue declining in the South region? / Which products should we discontinue?] **Data:** [Paste your tabular data here -- copied from Excel, Google Sheets, or as CSV] Analyze the data according to this framework: **1. Data Overview** - What do we have? (Number of data points, time period, categories) - Data quality: Are there missing values, inconsistencies, or obvious errors? **2. Key Findings (Top 5)** The 5 most important findings, ranked by relevance. For each: - What do the data show? (with specific numbers) - Why does this matter? - How confident is this finding? (clear / likely / hypothesis) **3. Trends and Developments** - What temporal trends are visible? - Are there seasonal patterns or cycles? - Where is a development accelerating or slowing? **4. Outliers and Anomalies** - Which data points deviate significantly from the average? - Are there sudden jumps or drops? - Possible explanations for the anomalies **5. Relationships and Correlations** - Which columns/variables are related? - Are there surprising correlations? - Important: Correlation does not equal causation -- label this clearly **6. Action Recommendations** 3-5 concrete, prioritized recommendations based on the data: - What should happen immediately? (Quick wins) - What should be addressed in the medium term? - What needs continued monitoring? **7. Recommended Visualizations** Which 3 charts would best illustrate the key findings? For each: - Chart type (line, bar, scatter, heatmap, etc.) - Which data on which axis - What the chart reveals Important rules: - Always cite specific numbers, not just 'has increased' - Clearly distinguish between fact (data shows X) and interpretation (this could mean Y) - If the data volume is too small for a reliable statement, say so honestly - Reference the given context if provided