DEV Community

ClawGear
ClawGear

Posted on

35 ChatGPT Prompts for Data Analysts (SQL Explanations, Stakeholder Reports, and Analysis Done Faster)

#chatgpt #dataanalysis #analytics #productivity

35 ChatGPT Prompts for Data Analysts (SQL Explanations, Stakeholder Reports, and Analysis Done Faster)

Data analysts spend a fraction of their time actually analyzing data. The rest goes to writing: explaining SQL queries, translating findings for non-technical stakeholders, drafting documentation, presenting insights in slide decks, and responding to ad-hoc requests that all feel urgent.

ChatGPT doesn't run your queries or replace your analytical judgment. But it will explain your query to a non-technical manager, turn a messy output table into a stakeholder-ready summary, write your data dictionary, and help you think through an analysis approach before you start. These 35 prompts are built for practicing data analysts — in-house, freelance, and agency — covering the communication and documentation work that buries your actual analysis time.

Section 1: Explaining Analysis to Non-Technical Stakeholders

Prompt 1 — Translate SQL to Plain English

Explain this SQL query in plain English for a non-technical stakeholder. Don't explain SQL concepts — just explain what the query does, what data it's pulling, what filters are applied, and what the result represents. [Paste SQL query]. Format as 3–4 sentences a VP of Marketing could read and immediately understand.
Enter fullscreen mode Exit fullscreen mode

Prompt 2 — Executive Summary of Analysis Findings

Write an executive summary of these data analysis findings for a C-suite audience. Key findings: [list 4–5 bullet points with numbers]. Keep it to 150 words. Lead with the most important insight, not the methodology. Use plain language — no technical terms. End with 1–2 recommended actions based on the data.
Enter fullscreen mode Exit fullscreen mode

Prompt 3 — Data Story Narrative

Turn these analysis results into a compelling data story narrative. Context: [what question we were answering]. Data: [key findings with numbers]. Audience: [non-technical business team]. Structure: Setup (the question), Conflict (what we found that was surprising or important), Resolution (what it means and what to do). 200–300 words, no jargon.
Enter fullscreen mode Exit fullscreen mode

Prompt 4 — Metric Explanation Email

Write an email to [stakeholder role: marketing team / sales leadership / product team] explaining what [metric name] means, how it's calculated, why it matters, and what the current value of [X] indicates about the business. Audience has no data background. Keep it under 200 words. Use one concrete analogy to make the metric tangible.
Enter fullscreen mode Exit fullscreen mode

Prompt 5 — Analysis Presentation Slide Notes

Write speaker notes for this slide in my data analysis presentation. Slide title: "[title]. Data shown: [describe chart/table]. Key takeaway I want the audience to leave with: [message]. Anticipated questions from the audience: [list 2]. Speaker notes should: expand on what the slide shows, pre-answer likely questions, and drive toward the key takeaway. 150–200 words."
Enter fullscreen mode Exit fullscreen mode

Section 2: Reports and Documentation

Prompt 6 — Weekly Metrics Report

Write a weekly metrics report for [team/stakeholder]. Key metrics this week: [paste numbers]. Compared to last week: [changes]. Compared to target: [on track / off track]. Write each section as: metric name, current value, change vs. prior period, 1-sentence interpretation. End with a "What to watch" section noting 1–2 things to monitor next week. Keep the whole report under 300 words.
Enter fullscreen mode Exit fullscreen mode

Prompt 7 — Data Dictionary Entry

Write a data dictionary entry for the field [field name] in the [table/dataset] table. Include: field name, data type, description (what it represents in plain English), example values, business logic or calculation (how it's derived), and any known limitations or gotchas. Audience: future analysts who will work with this data.
Enter fullscreen mode Exit fullscreen mode

Prompt 8 — Analysis Methodology Write-Up

Write the methodology section for an analysis on [topic]. What I did: [describe in plain terms — data sources, date range, filters applied, how I calculated the key metric, any assumptions I made]. Make it clear and replicable — another analyst should be able to reproduce the analysis from this description. Under 300 words.
Enter fullscreen mode Exit fullscreen mode

Prompt 9 — Data Quality Issue Report

Write a data quality issue report for [stakeholder / engineering team] about [issue: missing values / duplicate records / inconsistent formatting / broken data pipeline]. Include: issue description, impact (what analysis is affected and how), root cause (if known), current workaround (if applicable), and recommended fix. Factual, clear, no technical jargon beyond what the audience needs.
Enter fullscreen mode Exit fullscreen mode

Prompt 10 — Dashboard Documentation

Write documentation for a [dashboard name] dashboard used by [audience]. Include: purpose of the dashboard, key metrics shown and what each measures, data sources and refresh schedule, known limitations or caveats, and how to interpret the main chart/table. Audience: end users who didn't build it. Keep it conversational — no SQL, no technical implementation details.
Enter fullscreen mode Exit fullscreen mode

Section 3: Analysis Planning and Approach

Prompt 11 — Analysis Scoping Questions

I've been asked to analyze [business question / problem]. Before I begin, help me generate 10 scoping questions I should ask the stakeholder. Questions should cover: what decision will this analysis inform, what time period to examine, which customer/user segments to include, what "success" looks like, and what data is available. I want to avoid doing the wrong analysis.
Enter fullscreen mode Exit fullscreen mode

Prompt 12 — Analytical Framework Selection

I need to answer this business question: [question]. Suggest 3 analytical frameworks or approaches I could use to answer it. For each: describe the approach in plain English, what data it requires, what it would produce, and its limitation. Then recommend the best approach for my context: [describe available data and time constraints].
Enter fullscreen mode Exit fullscreen mode

Prompt 13 — Hypothesis Generation

I'm analyzing [business problem/metric: e.g., why user retention dropped 15% last quarter]. Generate 8 hypotheses for what might be causing this. For each hypothesis, suggest: what data would confirm or refute it, and which is most testable with available data ([describe data sources]). Organize from most to least likely based on common patterns in [industry].
Enter fullscreen mode Exit fullscreen mode

Prompt 14 — A/B Test Results Interpretation

Help me interpret these A/B test results: [paste results — control vs. variant, sample sizes, conversion rates, p-value or confidence interval]. Explain in plain English: (1) what the test showed, (2) whether the result is statistically significant and what that means, (3) what we should do next, and (4) any caveats or risks in the interpretation. Audience: non-technical product or marketing team.
Enter fullscreen mode Exit fullscreen mode

Prompt 15 — Root Cause Analysis Structure

Help me structure a root cause analysis for [observed problem: metric drop / data anomaly / business issue]. Use a 5-Whys or fishbone approach. Starting from: [the symptom]. Guide me through the layers, asking one "why" at a time, and generate a structured write-up I can present to stakeholders showing the logical chain from symptom to root cause.
Enter fullscreen mode Exit fullscreen mode

Section 4: Ad-Hoc Request Management

Prompt 16 — Clarifying Questions for Vague Requests

A stakeholder sent me this data request: "[paste vague request]". Before I start the analysis, generate 8 clarifying questions I should ask. Questions should resolve: the exact metric they need, the time period, the level of breakdown (by channel / region / segment), the format they want (table / chart / number), and the decision they're trying to make.
Enter fullscreen mode Exit fullscreen mode

Prompt 17 — Pushing Back on an Unrealistic Request

Write a professional email response to a stakeholder who asked for [request] by [unrealistic deadline]. I can realistically deliver [modified version] by [realistic date] instead. Explain: what's feasible in the timeframe, what tradeoffs they'd accept for a faster turnaround, and what I'd need from them to prioritize this over existing work. Collaborative, not defensive.
Enter fullscreen mode Exit fullscreen mode

Prompt 18 — Status Update on Analysis

Write a brief status update email to [stakeholder] on the [analysis name] analysis they requested. Current status: [where I am]. What's been completed: [list]. Remaining work: [list]. Expected delivery: [date]. Any blockers: [if applicable]. Keep it to 100 words. Factual and professional.
Enter fullscreen mode Exit fullscreen mode

Section 5: Data Visualization and Presentation

Prompt 19 — Chart Choice Explanation

I chose to visualize [metric/data] as a [chart type: bar chart / line chart / scatter plot / heatmap]. Write a 2-sentence annotation I can add to the chart or presentation slide explaining: (1) what to look at in this chart and (2) the key insight. The audience is non-technical. No mention of chart type — just guide their eyes to the point.
Enter fullscreen mode Exit fullscreen mode

Prompt 20 — Insight Annotation for a Chart

Write 3 annotation options for the most important finding in this chart: [describe what the chart shows and the key data point]. Each annotation should be under 20 words, call out the specific finding, and imply the action or implication. No hedging language. These will appear directly on the chart as callout boxes.
Enter fullscreen mode Exit fullscreen mode

Prompt 21 — Slide Rewrite (Data Slide)

Rewrite this data slide so it communicates insight, not just data. Current slide: [describe — shows X metric, Y timeframe, Z comparison]. The insight the slide should convey: [state the point]. Rewrite: headline (leading with the insight, not the metric name), subtitle (the supporting context), and 1-sentence annotation for the main chart element. Audience: VP-level, 30 seconds to absorb.
Enter fullscreen mode Exit fullscreen mode

Section 6: Stakeholder Communication

Prompt 22 — Findings Presentation Email

Write an email to [stakeholder] sharing the findings from the [analysis name] analysis. Key findings: [list 3–5 with numbers]. Recommendation: [what I recommend based on the data]. Next steps: [what I need from them or what happens next]. Attach: [describe what's attached]. Keep the email to 200 words — findings details are in the attached report.
Enter fullscreen mode Exit fullscreen mode

Prompt 23 — Pushing Back on a Data Misinterpretation

A senior stakeholder is drawing this conclusion from our data: "[their interpretation]". Based on the actual analysis, this conclusion is [incorrect / oversimplified] because [reason]. Write a tactful email or meeting follow-up that: acknowledges their interpretation, respectfully explains the limitation or nuance, provides the correct interpretation with data, and moves toward a productive next step. Under 200 words.
Enter fullscreen mode Exit fullscreen mode

Prompt 24 — Post-Analysis Retrospective

Write a brief post-analysis retrospective document (internal use only) for the [analysis name] project. Cover: what we set out to answer, what we found, what we'd do differently next time (data gaps, scope issues, process improvements), and any reusable components (queries, frameworks, templates) we should save. Audience: our analytics team. 300 words max.
Enter fullscreen mode Exit fullscreen mode

Section 7: SQL and Technical Communication

Prompt 25 — SQL Query Comment Block

Write a comment block for this SQL query that future analysts can read to understand it without running it. Include: purpose of the query, key tables and joins used, filters applied and why, output columns and what each represents, and any performance considerations or known issues. [Paste SQL here]. Keep it concise — under 200 words.
Enter fullscreen mode Exit fullscreen mode

Prompt 26 — Technical Explanation to Engineering

I need to request a new data field from the engineering team. The field I need: [description]. Why I need it: [analytical use case]. What table it should live in: [table]. Suggested data type and values: [description]. Frequency of updates needed: [real-time / daily / weekly]. Write a clear, concise engineering request in the format their team uses: requirements, not desires.
Enter fullscreen mode Exit fullscreen mode

Prompt 27 — Data Pipeline Documentation

Write documentation for a data pipeline that [describe what it does: pulls from X, transforms Y, loads into Z]. Include: overview, data sources and connections, transformation logic in plain English, output destination and format, schedule/trigger, owner, and known failure modes and how to handle them. For a new analyst who needs to maintain or troubleshoot this pipeline.
Enter fullscreen mode Exit fullscreen mode

Section 8: Career and Professional Development

Prompt 28 — LinkedIn Profile for a Data Analyst

Write a LinkedIn "About" section for a data analyst with [X] years of experience in [industry/domain]. Specialties: [tools and methods — SQL, Python, Tableau, etc.]. Notable projects or impact: [list 2–3]. Career goal: [stay IC / move to analytics manager / transition to data science / consulting]. Tone: technical credibility + human voice. 150–200 words.
Enter fullscreen mode Exit fullscreen mode

Prompt 29 — Technical Interview Prep

Help me prepare for a data analyst interview at [company type]. The role focuses on [domain: marketing analytics / product analytics / finance / operations]. Generate 10 behavioral and technical questions I'm likely to be asked, and for each: (1) write a model answer framework (STAR where applicable), and (2) note what the interviewer is actually looking for.
Enter fullscreen mode Exit fullscreen mode

Prompt 30 — Portfolio Project Write-Up

Write a portfolio project description for my personal website or GitHub. Project: [describe what you analyzed, what data you used, what tools, what you found]. Keep it accessible to a non-technical hiring manager and a technical engineer simultaneously. Structure: Problem, Data & Approach, Key Findings, Impact / What I Learned. Under 250 words.
Enter fullscreen mode Exit fullscreen mode

Prompt 31 — Performance Review Self-Assessment

Help me write my annual performance review self-assessment as a data analyst. My key contributions this year: [list]. Metrics or outcomes I can point to: [list]. A challenge I navigated: [describe]. Skills I grew: [list]. What I want to focus on next year: [goal]. Tone: confident, evidence-based, growth-oriented. 400–500 words.
Enter fullscreen mode Exit fullscreen mode

Section 9: Specialized Analysis Communication

Prompt 32 — Cohort Analysis Explanation

Explain the results of this cohort analysis to a non-technical marketing team: [describe results — cohorts defined by X, metric tracked is Y, key finding is Z]. Use an analogy that makes cohort thinking intuitive. Then summarize: what the data shows about [customer behavior], what it implies for [marketing / product / retention strategy], and what we should do next.
Enter fullscreen mode Exit fullscreen mode

Prompt 33 — Churn Analysis Summary

Write a churn analysis summary for [audience: exec team / product team / customer success]. Analysis covered: [time period, customer segment, churn definition]. Key findings: [paste]. Write as: What we looked at, What we found (top 3 insights with numbers), What drives churn (ranked factors), and Recommended actions (3 bullet points). Under 300 words.
Enter fullscreen mode Exit fullscreen mode

Prompt 34 — Forecasting Assumptions Document

Write the assumptions document for a [revenue / demand / user growth] forecast I built. Methodology: [describe]. Key assumptions: [list]. Uncertainty range: [describe]. Scenarios modeled: [base / upside / downside]. What would cause the forecast to be wrong: [list risks]. For a leadership team that will use this forecast for planning. Clear, honest, under 300 words.
Enter fullscreen mode Exit fullscreen mode

Prompt 35 — "What Does This Number Mean?" Response

A stakeholder asked: "What does [metric: e.g., our 34% 30-day retention rate] actually mean? Is that good or bad?" Write a response that: defines the metric in plain English, provides industry benchmark context, interprets our current value (what it means for the business), and suggests 1 action to improve it. Under 150 words. Concrete, not vague.
Enter fullscreen mode Exit fullscreen mode

Make Every Analysis Count

These 35 prompts cover the communication layer that determines whether your analysis drives decisions or sits in a shared drive unopened. The best analysts aren't just technically strong — they translate data into language that moves people.

Want 50 more data analyst prompts? The Data Analyst AI Toolkit includes advanced prompts for exploratory data analysis write-ups, stakeholder relationship management, analytics strategy documents, and building your analytics career. ($14.99, instant download.)

More professional ChatGPT prompt guides: copywriters | sales reps | customer success managers

Top comments (0)