AI Data Analysis Cheatsheet

Data Exploration

"Here is my dataset [paste or upload CSV]. Describe the structure: how many rows, columns, data types, and any obvious data quality issues."

"Calculate summary statistics for each numeric column: mean, median, min, max, standard deviation, and percentage of missing values."

"What are the most interesting patterns or anomalies in this data? Start with the most surprising finding."

"Identify correlations between these columns: [list columns]. Which relationships are strongest?"

"Are there any outliers in [column]? List them and suggest whether they are errors or legitimate extreme values."

Data Cleaning Prompts

"Identify all data quality issues in this dataset: missing values, duplicates, inconsistent formats, and obvious errors."

"Write a Python/SQL script to clean this dataset: standardize date formats, remove duplicates, and fill missing values with [median / mode / forward-fill]."

"These columns have inconsistent naming [paste examples]. Create a mapping table to standardize them."

"Find and flag rows where [column A] and [column B] have logically impossible combinations."

"Deduplicate these records. Suggest which record to keep when duplicates have conflicting values."

Analysis and Insights

"Segment this customer data into meaningful groups. Describe each segment and suggest targeted strategies for each."

"Compare performance across [time periods / regions / categories]. Which segments are improving and which are declining?"

"What factors most strongly predict [outcome variable]? Rank them by importance."

"Perform a cohort analysis on this user data. How does retention change across monthly signup cohorts?"

"Run a year-over-year comparison for [metric]. Highlight months with significant changes and suggest explanations."

"Calculate the lifetime value of customers in each [segment / channel / plan tier] from this transaction data."

Visualization Requests

"Suggest the best chart type to visualize [this data / this relationship]. Explain why."

"Write Python code (matplotlib/seaborn) to create a [bar chart / line graph / heatmap / scatter plot] showing [what to visualize]."

"Create a dashboard layout for this data. What 4-6 charts would give a complete picture of [business area]?"

"Generate a chart comparing [metric] across [categories] with a trend line showing the direction over time."

"Create a before-and-after visualization showing the impact of [change/initiative] on [metric]."

Reporting and Communication

"Write a 3-paragraph executive summary of these findings. Lead with the most impactful insight."

"Turn this data analysis into 5 bullet points for a slide deck. Each bullet should have a stat and a plain-English interpretation."

"What questions would a skeptical executive ask about these findings? Prepare answers for each."

"Translate these technical analysis results into recommendations a non-technical stakeholder can act on."

"Create a one-page data brief covering: key finding, supporting evidence, implications, and recommended next steps."

SQL and Spreadsheet Help

"Write a SQL query to [describe what you need] from a table with columns: [list columns]."

"Convert this SQL query into plain English so I can verify it does what I expect: [paste query]."

"Write an Excel formula to [calculate X] given data in columns [A, B, C]."

"Create a pivot table structure to analyze [metric] by [dimension 1] and [dimension 2]."

"Optimize this slow SQL query. Explain what makes it slow and how your version is faster: [paste query]."

Analyze data faster without being a data scientist. These prompt templates help you explore datasets, find patterns, create visualizations, and generate insights using AI — whether your data lives in spreadsheets, CSVs, or databases.