Data Analysis Prompts

You are a CRM analyst writing {{sql_dialect}}. Transactions table: {{transactions_table}} (customer_id, order_date, order_value). Analysis as-of date: {{as_of_date}}. Number of sco...

Act as a data visualization expert following Cleveland and Few principles. The message I want the viewer to take away: {{key_message}} Data I have: {{data_description}} Number of c...

You are a Python data viz engineer who writes clean Matplotlib. Chart type: {{chart_type}} Dataframe columns to plot: {{columns}} The takeaway the chart must convey: {{takeaway}} B...

Act as an analytics lead designing a dashboard. Business goal the dashboard supports: {{business_goal}} Primary audience and how often they look: {{audience_cadence}} Decisions thi...

You are a spreadsheet power user for {{spreadsheet_app}}. What I want the formula to do: {{goal}} Data layout (sheet, ranges, headers): {{layout}} Edge cases to handle: {{edge_case...

Act as a spreadsheet analyst. My raw data columns: {{columns}} The question I need answered: {{question}} Tool: {{spreadsheet_app}} Task: 1. Tell me exactly what to put in Rows, Co...

You are a skeptical causal-inference reviewer. The claim being made: {{claim}} Evidence behind it: {{evidence}} How the data was collected: {{data_collection}} Proposed action if t...

Act as an analyst who writes crisp insight summaries. Here is the results table: {{results_table}} Context and the question this addresses: {{context}} Audience: {{audience}} Task:...

You are a data steward documenting a table. Table name: {{table_name}} Columns with sample values: {{columns_with_samples}} What the table represents: {{table_purpose}} Grain: {{gr...

Act as a SQL reviewer who explains queries clearly. Query ({{sql_dialect}}): {{query}} Task: 1. Summarize in one sentence what this query returns. 2. Walk through it step by step,...

You are an analyst running a root-cause investigation. Metric that moved: {{metric}} What happened: {{change_description}} between {{period_a}} and {{period_b}} Dimensions I can sl...

Act as a data quality engineer. Table / dataset: {{dataset}} Columns and meanings: {{columns}} How the data is used downstream: {{downstream_use}} Known past failures: {{past_failu...

You are a SQL expert in {{sql_dialect}}. Table: {{table_name}} with {{entity_column}}, {{sequence_column}}, and {{status_column}}. I want to find: {{goal}} (for example consecutive...

Act as an applied statistician. My question / hypothesis: {{hypothesis}} Outcome variable type: {{outcome_type}} Grouping or predictor variable(s): {{predictors}} Sample size: {{sa...

You are a quantitative marketing analyst. Business model: {{business_model}} (subscription, transactional, freemium) Data available: {{data_available}} Time horizon for LTV: {{hori...

You are a data analyst profiling an unfamiliar table in {{sql_dialect}}. Table: {{table_name}} Columns: {{columns}} Task: write SQL that profiles the table and returns, per relevan...

Act as a time series analyst. Series: {{series_description}} Frequency: {{frequency}} Suspected seasonal periods: {{seasonal_periods}} Task: 1. Recommend additive vs multiplicative...

You are a marketing analytics specialist. Conversion event: {{conversion_event}} Touchpoint data available: {{touchpoint_data}} Typical path length: {{path_length}} Goal of the att...

Act as a survey methodologist. Survey questions and scales: {{questions_and_scales}} Number of respondents: {{n_respondents}} Key comparisons I want: {{comparisons}} Segments: {{se...

You are a data scientist scoping a churn model. Business: {{business_description}} How churn is defined here: {{churn_definition}} Data available: {{data_available}} What the predi...

You are a Power BI developer fluent in DAX. Data model: {{tables_and_relationships}} The measure I need: {{measure_goal}} Filter context it must respect or override: {{filter_conte...

Act as a statistician communicating uncertainty. Estimate type: {{estimate_type}} (mean, proportion, difference) Observed value: {{observed_value}} Sample size: {{sample_size}} Var...

You are a data engineer working in {{sql_dialect}}. Column {{json_column}} in table {{table_name}} contains JSON shaped like: {{json_sample}} I need a flat table with these fields:...

Act as a data scientist running customer or item segmentation. Entities to cluster: {{entities}} Available features: {{features}} Goal of the segmentation: {{goal}} Number of clust...

You are a spreadsheet analyst for {{spreadsheet_app}}. My data: {{data_layout}} (one row per user with signup date and activity dates or a long event log). Retention period unit: {...

Act as a pricing analyst. Product: {{product}} Historical data available: {{data_available}} (prices, quantities, dates, promotions) Confounders to worry about: {{confounders}} Tas...

You are a SQL expert in {{sql_dialect}}. Table: {{table_name}} with id {{id_column}} and parent {{parent_column}}. What I need: {{goal}} (for example full path to root, depth level...

Act as a statistician advising on missing data. Columns with missingness and their null rates: {{columns_and_rates}} What I know about why values are missing: {{missingness_reason}...

You are an analytics lead reviewing a proposed metric. Metric name: {{metric_name}} Proposed definition: {{proposed_definition}} What decision it informs: {{decision}} Data source:...

Act as a Bayesian experimentation analyst. Metric type: {{metric_type}} Control data: {{control_data}} Variant data: {{variant_data}} Prior beliefs (or use weak priors): {{prior}}...

You are an analytics engineer using {{sql_dialect}} and {{transformation_tool}}. Source table: {{source_table}} (high volume, append-mostly) Target model grain: {{target_grain}} Up...

Act as a geospatial data analyst. Location data: {{location_data}} (points, regions, or both) Question to answer: {{question}} Tools available: {{tools}} (PostGIS, GeoPandas, BigQu...

You are an ML practitioner documenting a model responsibly. Model purpose: {{model_purpose}} Algorithm and version: {{algorithm}} Training data: {{training_data}} Performance metri...

Act as an analyst building a composite index. What the score should measure: {{score_purpose}} Input metrics and their directions (higher better or worse): {{input_metrics}} How we...

You are an analyst presenting findings to {{audience}}. The analysis I did: {{analysis_summary}} Key findings: {{key_findings}} The decision I want them to make: {{desired_decision...

You are a data analyst doing reconciliation in {{sql_dialect}}. Source A: {{source_a}} Source B: {{source_b}} Join key: {{join_key}} They should agree on: {{agreement_fields}} Task...

Act as a data analyst who distrusts averages. Variable: {{variable}} Context: {{context}} The question stakeholders asked using the average: {{average_question}} Task: 1. Explain w...

You are an ML interpretability specialist. Model type: {{model_type}} Top features and their importance scores or SHAP summary: {{importance_data}} Audience: {{audience}} What deci...

Act as a hostile but fair peer reviewer of my analysis. The analysis and its conclusion: {{analysis_and_conclusion}} Method and data used: {{method_and_data}} The decision riding o...

You are an analytics engineer building self-service query templates in {{sql_dialect}}. The recurring questions the team asks: {{recurring_questions}} Key tables and their grain: {...

You are a SQL engineer working in {{sql_dialect}}. Table: {{table_name}} Timestamp column: {{timestamp_column}} Expected interval: {{interval}} (for example hourly or daily) Option...

Act as a forensic data analyst. I have a column of {{amount_column}} values from {{dataset_description}}. Walk me through a Benford's Law first-digit analysis: 1. State the expecte...

You are a database performance engineer for {{database_engine}}. Here is my query: {{query}} Here is the EXPLAIN (or EXPLAIN ANALYZE) output: {{explain_output}} Do the following: 1...

You are an analytics engineer. Source tables: - Users: {{users_table}} with id and signup timestamp {{signup_column}}. - Events: {{events_table}} with user id and event timestamp {...

Act as a data quality specialist. I need to deduplicate {{record_type}} records with columns: {{key_columns}}. The data is dirty: typos, casing, abbreviations, and reordered tokens...

You are a time-series analyst. Metric: {{metric_name}}, observed at {{frequency}} frequency over {{time_span}}. Guide me through a decomposition: 1. Recommend additive vs multiplic...

You are a data visualization designer. I have path data describing how users move between stages: {{stage_list}}. Each row has source stage, target stage, and a count. Help me buil...

You are a spreadsheet power user. Platform: {{spreadsheet_app}} (Google Sheets or Excel). Column {{source_cell}} contains text like: {{sample_values}} I want to extract: {{target_p...

Act as an experimentation statistician who refuses to overclaim. Experiment: {{experiment_name}} Metric: {{primary_metric}} Control: {{control_n}} users, {{control_value}} Variant:...

You are a data wrangler. Current shape of {{table_name}}: {{current_layout}} Target shape: {{target_shape}} (long or wide). The identifier columns are {{id_columns}} and the value...