AI Prompts for Gemini for Data Analysis

Here is a description of my dataset: [DESCRIBE COLUMNS, ROW COUNT, SOURCE]. Help me assess: what analyses are possible with this data, what it cannot tell me, any obvious quality issues I should fix before analysis, and what additional data would significantly improve what I can learn.

I have a dataset with these quality issues: [DESCRIBE ISSUES, e.g. missing values in X column, duplicate rows, inconsistent date formats]. Write the specific cleaning steps and code [IN PYTHON/SQL/SHEETS FORMULA] to fix each issue. Explain the trade-offs for each approach, especially how to handle missing values.

I have a Google Sheet with this structure: [DESCRIBE COLUMNS AND LAYOUT]. I need to: [DESCRIBE WHAT YOU WANT TO DO, e.g. calculate rolling averages, flag duplicates, join two sheets on a key column]. Write the specific formulas, where to put them, and how they work.

Write a SQL query to extract [DESCRIBE WHAT YOU NEED] from this database schema: [DESCRIBE TABLES AND COLUMNS]. The query should: [LIST REQUIREMENTS, e.g. filter by date range, aggregate by category, join on customer_id]. Optimize for readability and include comments explaining the logic.

Read this sample of my dataset: [PASTE SAMPLE OR DESCRIBE STRUCTURE]. Identify any data quality issues: outliers that might be errors, impossible values, inconsistent formatting, referential integrity problems, or missing data patterns. For each issue, tell me how to investigate it and how to handle it in analysis.

I want to answer this business question: [DESCRIBE QUESTION] using this data: [DESCRIBE DATA]. What is the most appropriate analysis method? Give me two or three options with their trade-offs. For the best option, describe what output it produces and what I need to interpret it correctly.

Write a Python script to analyze [DESCRIBE DATASET AND QUESTION]. Use pandas and appropriate visualization libraries. The script should: load the data from [FILE FORMAT], perform [SPECIFIC ANALYSES], produce [SPECIFIC OUTPUTS], and save results to [OUTPUT FORMAT]. Include error handling and comments.

I have data in Google Sheets: [DESCRIBE STRUCTURE]. I want to: [DESCRIBE ANALYSIS]. Walk me through doing this analysis entirely in Sheets, using built-in functions or a Gemini-assisted approach. Provide the exact formulas or steps, and tell me how to check that the results are correct.

Write a BigQuery SQL analysis for [DESCRIBE BUSINESS QUESTION] using this table structure: [DESCRIBE SCHEMA]. Include: the main analysis query, any intermediate CTEs needed for clarity, and a brief interpretation of what each step produces. Optimize for cost by minimizing data scanned.

I want to run a [COHORT/SEGMENTATION/FUNNEL] analysis on this data: [DESCRIBE DATA]. Write the analysis code in [PYTHON/SQL] that produces: [DESCRIBE OUTPUTS, e.g. retention by cohort, conversion by segment, drop-off by funnel stage]. Include the visualization code.

Here are the results of my analysis: [PASTE RESULTS OR DESCRIBE]. The business question I was trying to answer: [DESCRIBE]. Interpret these results: what do they actually mean in plain language, what is the most important finding, what limitations of the analysis should I flag, and what does this imply I should do?

I have data showing [METRIC] changed by [AMOUNT] over [PERIOD]. Here is the data: [PASTE OR DESCRIBE]. Help me identify the key drivers of this change. What analysis would decompose this change into contributing factors? Write the analysis code and interpret the output.

Read this data and identify any anomalies, outliers, or unexpected patterns: [PASTE DATA OR RESULTS]. For each one, tell me: what makes it unusual, whether it is more likely to be a real finding or a data error, and what I should investigate to understand it.

I ran an A/B test with these results: [DESCRIBE RESULTS, sample sizes, conversion rates or metric values for control and variant]. Help me: calculate statistical significance, determine if the result is practically meaningful given sample size, identify any threats to validity in my test design, and decide whether to ship the variant.

Here are my data findings: [DESCRIBE FINDINGS]. My stakeholder audience is [DESCRIBE]. What are the three most important business implications of these findings? What actions should we take, what should we investigate further, and what should we stop doing? Be direct and specific.

Write a data analysis summary for [AUDIENCE, e.g. executive team / product team / investors]. Key findings: [LIST]. Include: the business question, the key metrics, the most important finding, the recommended action, and the key caveats. Under 300 words.

I want to build a Looker Studio dashboard tracking [BUSINESS AREA] for [AUDIENCE]. Based on my data structure: [DESCRIBE], recommend: the five most important metrics to track, the best visualization type for each, how to arrange the layout for a daily business review, and what filters to include.

I am presenting these data findings: [DESCRIBE] to [AUDIENCE]. Write five data-driven talking points I can use in the presentation. Each should: cite a specific number, explain what it means in context, and connect to a business implication. Write them as spoken sentences, not bullet points.

I have these data findings: [DESCRIBE]. Help me structure them as a data story for a presentation. Use the Problem-Evidence-Insight-Action structure: (1) the business problem that motivated the analysis, (2) the key evidence from the data, (3) the insight that emerges, (4) the recommended action. Write the narrative arc.

Write documentation for this analysis so another analyst can reproduce it: [DESCRIBE ANALYSIS AND CODE]. Include: the business question, the data sources and access details, the processing steps, the analysis methodology, the interpretation logic, and known limitations. Format as a shareable analysis README.