AI Prompts for Grok for Data Analysis

I have data about [DESCRIBE DATASET]. The business question I need to answer is: [DESCRIBE QUESTION]. Help me clarify the exact analysis question: what specifically am I measuring, what comparison or change will answer the question, and what data do I actually need versus what I have? Flag any assumptions in my question that I should test.

My question is: [ANALYSIS QUESTION]. My data includes: [DESCRIBE COLUMNS AND TYPES]. What type of analysis is most appropriate: descriptive statistics, trend analysis, cohort analysis, A/B test evaluation, regression, or something else? What are the key assumptions this approach requires that I should check first?

My [METRIC NAME] is currently [VALUE]. Is this good, bad, or average for [INDUSTRY/BUSINESS TYPE]? What are the current industry benchmarks? Has this benchmark changed recently? What factors typically explain variance above or below benchmark in businesses like mine?

I want to answer the question: [QUESTION]. My current data includes: [DESCRIBE]. What data am I missing that would make this analysis more reliable or complete? For each gap, tell me: how important it is, where I might be able to get it, and what I can do in the meantime if I cannot get it.

I need to analyze [COMPLEX TOPIC] to inform a decision about [BUSINESS DECISION]. Break this into a logical sequence of analysis steps. What do I need to understand first before I can answer the harder questions? What is the right order to build up from simple to complex? What are the decision points where earlier results will determine which analysis to run next?

Write a SQL query that answers: [QUESTION]. My relevant tables are: [DESCRIBE SCHEMA]. Conditions: [DESCRIBE FILTERS, JOINS, AGGREGATIONS]. Return: [DESCRIBE DESIRED OUTPUT COLUMNS]. Optimize for: [DESCRIBE CONSTRAINTS — SIZE OF TABLE, INDEX ON X]. Add a comment explaining any non-obvious logic.

Write a Python script using pandas to: [DESCRIBE ANALYSIS TASK]. My data is in [CSV / DATAFRAME / FROM SQL QUERY]. The data has these columns: [LIST COLUMNS]. Steps needed: [DESCRIBE STEPS]. Output should be: [DESCRIBE OUTPUT — SUMMARY TABLE / CHART / SAVED FILE]. Include error handling for missing values.

I need to aggregate [TABLE/DATASET] by [GROUPING DIMENSIONS] and calculate [METRICS — COUNTS, AVERAGES, PERCENTILES]. The result should show [DESCRIBE DESIRED SHAPE — ONE ROW PER X, COLUMNS FOR EACH Y]. Write this in [SQL / PYTHON PANDAS / DPLYR]. Include a version that handles nulls appropriately.

I need to transform raw data from [SOURCE FORMAT] to [TARGET FORMAT]. Transformations needed: [DESCRIBE — RESHAPE WIDE TO LONG, JOIN TWO TABLES, CALCULATE DERIVED COLUMNS, FILTER ROWS BY CONDITION, FILL MISSING VALUES WITH X]. Write the transformation in [SQL / PYTHON / R]. Show the before and after schema.

This query is supposed to [DESCRIBE EXPECTED RESULT] but returns [DESCRIBE ACTUAL RESULT]: [PASTE QUERY]. Walk through what this query is actually doing step by step. Identify where the logic diverges from my intention and show me the corrected version.

I ran [DESCRIBE ANALYSIS] and got: [PASTE RESULTS]. My sample size is [N]. What does this result mean in plain English? Is it statistically significant? Is the effect size meaningful for a business decision? What are the most important caveats on what I can conclude?

I have this analysis result: [DESCRIBE FINDING]. Before I present this, help me check: could the result be an artifact of data quality issues (missing data, outliers, biased sampling)? What would I need to check to be confident the result is real? Are there any obvious ways this analysis could be wrong?

My data shows [CORRELATION OR RELATIONSHIP — E.G., CUSTOMERS WHO DO X HAVE 40% HIGHER RETENTION]. Before I present this as causal, what are the three most plausible alternative explanations? How could I test whether the relationship is causal or just correlational with the data I have?

I segmented my data by [SEGMENTATION CRITERIA] and found [DESCRIBE DIFFERENCES BETWEEN SEGMENTS]. Is this segmentation meaningful? Are the groups large enough? Is the difference between groups statistically significant? Could the segmentation be creating an artifact that does not reflect real underlying differences?

My analysis shows: [DESCRIBE FINDINGS]. Before I use this to make a decision about [DECISION], help me identify: what the data CANNOT tell me, what additional data I would need to be fully confident, and what assumption I am implicitly making when I use this data to justify the decision.

My analysis found: [LIST KEY FINDINGS]. The business question was: [QUESTION]. Write a three-paragraph executive summary for [AUDIENCE — EXECUTIVES / BOARD / PRODUCT TEAM]. Paragraph one: what question we were answering and the key finding. Paragraph two: three most important supporting data points. Paragraph three: the recommended action and next steps.

I need to explain this finding to a non-technical audience: [DESCRIBE FINDING]. Write an explanation in plain English with: an analogy that makes the concept immediately clear, the specific business implication (so what), and what action this finding suggests. Maximum two sentences plus the analogy.

I have these findings to visualize: [LIST FINDINGS]. For each one, recommend the most effective chart type: bar, line, scatter, heatmap, funnel, histogram, or other. Explain why that chart type best communicates each specific insight and flag any common visualization mistakes I should avoid for that type of data.

My analysis shows: [DESCRIBE FINDINGS]. The decision we need to make is: [DESCRIBE DECISION]. Help me write a one-page recommendation that: states the recommended action clearly, cites the three strongest data points that support it, acknowledges the main uncertainty or risk, and defines what we will watch to know if the decision was right.

I am presenting analysis about [TOPIC] to [AUDIENCE]. My key findings are: [LIST]. Build a narrative structure for the presentation: the setup (why this analysis matters now), the complication (what the data reveals that is unexpected), the resolution (what we should do), and the call to action. Map each finding to a slide in this structure.