Prompts that brief agents, not chatbots
Copy, paste, then add your files and stop conditions. Start with Agentic if you use ChatGPT, Claude, Gemini, Copilot, Cursor, or Claude Code.
Prompt of the Day
A different prompt every day. Copy, paste, and put it to work.
Saved prompts: 0
Work starters
Copy, fill the brackets, check before you share. Works in Copilot, ChatGPT, Claude, or Gemini.
Work
P1. Status update from notes
Turn these notes into a status update for [AUDIENCE]. Format: 3 bullets (done, next, blocked) and one ask. Under 120 words. Use only what is in the notes. Mark anything unclear with [CHECK]. Notes: [PASTE]
Check before sharing: Every date, name, and number matches your notes.
Work
P2. Variance explanation
Here is plan vs actual for [METRIC]. Explain the variance in plain language. Give the top 3 drivers with the amount each contributes, and label each driver as confirmed from the data or a hypothesis. Do not invent drivers. Data: [PASTE OR ATTACH]
Check before sharing: Drivers add up to the total variance in the source.
Work
P3. Meeting prep from email
Prep me for [MEETING] with [PEOPLE]. From my recent emails and files, list: what was decided, open questions, what they will likely ask me, and 3 points I should make. Link each item to its source.
Check before sharing: Open the linked sources; drop anything stale or from the wrong project.
Work
P4. Meeting recap to actions
From this transcript or notes, list decisions, action items (owner, due date), and open questions. If an owner or date was not stated, write 'not stated.' Notes: [PASTE]
Check before sharing: Each owner agreed to their item in the meeting.
Work
P5. Long thread to reply
Summarize this thread in 3 lines, then draft my reply. Tone: [DIRECT / WARM]. Answer only what was asked of me. Keep it under 100 words. Thread: [PASTE]
Check before sharing: The reply commits to nothing you have not approved.
Work
P6. One-page decision memo
Write a one-page decision memo. Sections: decision needed, options (2-3), trade-offs, recommendation, risks, deadline. Use only these facts: [PASTE]. Flag any missing fact instead of guessing.
Check before sharing: The recommendation is yours, and every fact has a source.
Work
P7. Explain this spreadsheet
Explain what this sheet does, tab by tab. List every formula that references another tab or file. Flag hard-coded numbers inside formulas and any tabs you could not read. [ATTACH]
Check before sharing: Spot-check two totals by hand.
Work
P8. Rewrite for a senior reader
Rewrite this for [ROLE] who has 60 seconds. Lead with the answer, then 3 supporting points, then the ask. Keep all numbers exactly as written. Draft: [PASTE]
Check before sharing: No number or claim changed meaning.
Work
P9. What did I miss?
From my mail, chats, and meetings in the last [N] days, list what needs my reply or decision, sorted by urgency. For each: who, what they need, by when, and a link.
Check before sharing: Open the top three items yourself before acting.
Work
P10. Process to checklist
Turn this process description into a numbered checklist. Add who does each step, what 'done' looks like, and where a human must approve. Mark unclear steps [ASK]. Process: [PASTE]
Check before sharing: Walk through it once with someone who does the task.
Work
P11. Critique my draft
Act as a skeptical reviewer. List the 5 weakest points in this draft: unsupported claims, missing data, unclear asks. For each, suggest a fix. Do not rewrite the whole thing. Draft: [PASTE]
Check before sharing: You decide which critiques to accept.
Work
P12. Proof receipt
Turn this log into a proof receipt: task, tool, before, after, human check, outcome. Then write one resume line and three talking points for my manager. Use only numbers in the log and label each as measured or estimated. Log: [PASTE]
Check before sharing: Every number traces to your log.
Team lead · Monday decision
Use case / one slide, not a dump
Agent brief recipe
The brief is the skill. Six fields. Then a three-row decision — Risk, Impact, Ask, Owner.
Do not invent status. If a fact is missing, mark it Unknown. A lead approves every fact and owner before anyone sends.
- Goal
- One decision slide for Monday — three rows, no narrative dump.
- Context
- Last week’s notes and the owner list (attached). No extra hunting.
- Tools
- Read files. Draft the table. Do not send.
- Constraints
- Do not invent status. If a source is missing, write Unknown — never fill a cell to look complete.
- Approval gate
- Lead approves every fact and owner before send.
- Return format
- Table: Risk · Impact · Ask · Owner. Maximum three rows.
How / decision preview
Example · 3 rows| Risk | Impact | Ask | Owner |
|---|---|---|---|
| Vendor slip on the Friday extract | Close pack late by one day | Hold the 10:00 review? | Jordan |
| License true-up unconfirmed | $6k swing vs plan | Confirm the source tab | Priya |
| Overtime coding incomplete | Driver may be overstated | Re-cut after payroll | Unknown |
Human gate Lead approves every fact/owner before send.
Agentic prompts
Use these in ChatGPT with tools, Claude with computer use, Copilot agents, Cursor, or Claude Code. The brief is the skill.
Brief
Run this as an agent
You are an agent, not a chatbot. Goal: [TASK]. Context attached: [FILES OR NOTES]. Tools you may use: [SEARCH / CODE / BROWSER / NONE]. Constraints: do not invent numbers; stop and ask if a source is missing. Loop: plan, act, check. Return (1) the deliverable, (2) what you did, (3) what a human must verify before this ships.
Coding
Brief a coding agent (Cursor or Claude Code)
Repo: [PATH OR REPO NAME]. Task: [BUG / FEATURE / REFACTOR]. Do not rewrite unrelated files. Before editing, list the files you will touch and why. After editing, run [TEST COMMAND] and paste the output. If tests fail, fix once, then stop and report. I will review the diff before anything is committed.
Team
Researcher, drafter, critic
Run three roles in order on [TASK]. Researcher: gather only sourced facts, list unknowns. Drafter: write the deliverable from those facts only. Critic: mark every unsupported claim, rate risk High/Medium/Low, and say what I must check. Do not merge the roles. Show each role's output under its own heading.
Tools
MCP or connected-tool agent
You can use these connected tools: [LIST: calendar, files, CRM, Slack, none]. Task: [TASK]. Before each tool call, say why. After each call, say what you learned in one sentence. Never send, delete, or write to a live system without my explicit yes. End with a log of tool calls and a draft I can approve.
Browser
Computer-use / browser agent
Do this UI task: [STEPS, e.g. download last week's report from [URL], extract the three KPIs, put them in a table]. Stay inside [SITE]. If a login, captcha, or unexpected page appears, stop and tell me. Do not click Purchase, Submit, or Delete. Return a screenshot description plus the extracted table.
QA
Eval this agent on five cases
I will paste an agent brief and five test cases. For each case, run the brief, then score: did it follow tools limits, did it invent facts, did it stop when it should. Table: case, pass/fail, failure mode, one fix to the brief. Do not rewrite the whole prompt unless two or more cases fail the same way. Brief: [PASTE]. Cases: [PASTE].
Review
Draft, then attack your own draft
Task: [DELIVERABLE]. First write it. Then, without me asking, switch to reviewer: list claims without sources, weakest paragraph, and three questions a skeptical manager would ask. Revise once. Show original issues and the revised draft. If you cannot source a claim, cut it.
Gate
Work until the approval gate
Goal: [OUTCOME]. You may read, search, and draft. You may not send email, change production data, or publish. When you reach a point that needs my approval, stop with: proposed action, risk, and a yes/no question. Wait. If I say no, offer one alternative. Context: [PASTE].
AI Agents Prompts
Shorter multi-step jobs: research with sources, a spec, a campaign, an audit.
Research
Research and Summarize with Sources
Research [TOPIC] and summarize the key findings. Include: 3-5 main points, sources with URLs, and 3 actionable recommendations. Use web search. End with a short list of follow-up questions to validate the findings. Best in ChatGPT with Browse or Gemini with Google Search.
Analytics
Analyze Data and Produce Exec Summary
I will paste data below. Analyze it in 3 steps: 1) List observed patterns only (no guesses). 2) List hypotheses separately. 3) Produce a one-page exec summary with: what changed, impact, and 3 actions. Format as bullets under 15 words each. Do not invent numbers. Paste data: [PASTE HERE]
Marketing
Campaign Brief to Assets to Test Plan
Create a mini campaign for [PRODUCT] targeting [AUDIENCE]. Step 1: One-line offer, 3 persona angles (pain, outcome, objection). Step 2: 4 ad hooks, 3 email subject lines, 1 landing hero. Step 3: 2-week A/B test plan with hypothesis, KPI, sample target. Keep claims specific and compliant. Best in ChatGPT or Claude with iterative follow-ups.
Product
Feature Request to Sprint-Ready Spec
Convert this feature request into a sprint-ready spec. Request: [PASTE REQUEST]. Output: A) One-sentence problem. B) In-scope / out-of-scope. C) 5 acceptance criteria in Given/When/Then. D) Top 5 edge cases. E) Telemetry events to track. Then create two implementation options: Lean (2 weeks) vs Robust (6 weeks) with tradeoffs. Best in Claude or ChatGPT.
QA
Audit Your Own Analysis
Audit this analysis for errors. Check: denominator changes, seasonality, outliers, small sample bias, channel mix shifts. For each risk, mark High/Medium/Low and suggest one validation query. Do not guess at causes — list what to verify. Paste analysis: [PASTE HERE]
Competitive
Competitive Intelligence Brief
Research [COMPANY/INDUSTRY] and produce a competitive brief. Use web search. Include: positioning, recent news, pricing signals, product updates, marketing channels, and 3 opportunities or threats for [OUR COMPANY]. Cite sources. End with 5 questions a sales or product team should ask to validate. Best in ChatGPT with Browse or Perplexity.
Marketing Prompts
Strategic marketing, campaign planning, and brand positioning prompts.
Tactical
Message Map to Launch Assets (30-Min Workflow)
Product: [PRODUCT]. Audience: [AUDIENCE]. Goal: [LEADS/DEMO/TRIAL]. Channel: [LINKEDIN/EMAIL/LANDING]. Step 1: Return a message map with 1 core promise, 3 proof points, 3 objections + responses. Step 2: Using that map, generate 5 ad hooks, 3 email subject lines, 2 CTA variants. Step 3: Create a 2-week A/B test plan with hypothesis, KPI, and stop/scale rules. No hype, max character limits included.
Tactical
Audience Angles for [PRODUCT]
For [PRODUCT] targeting [AUDIENCE], give 3 persona angles: Pain, Desired Outcome, Objection. Then create 4 LinkedIn ads, 3 email subject lines, 1 landing hero section. Keep claims compliant and specific. Avoid "revolutionary" and "game-changing." Use grade-8 reading level. CTA must start with an action verb.
Campaign Strategy
Create a Marketing Campaign Brief
Act as a senior marketing strategist. Create a comprehensive campaign brief for [PRODUCT/SERVICE] targeting [AUDIENCE]. Include: campaign objectives, key messages, channel strategy, timeline, KPIs, and budget allocation recommendations. The campaign should differentiate us from [COMPETITORS] and leverage our unique value proposition of [UVP].
Competitive Analysis
Competitive Landscape Analysis
Analyze the competitive landscape for [INDUSTRY/MARKET]. Create a framework comparing: positioning, pricing strategies, target audiences, marketing channels used, content strategies, and unique differentiators. Identify gaps and opportunities for a new entrant or existing player to capture market share.
Audience Persona
Detailed Buyer Persona Creation
Create a detailed buyer persona for [PRODUCT/SERVICE]. Include: demographic details, psychographic profile, pain points, goals, objections to purchase, preferred content formats, media consumption habits, buying triggers, and a day-in-the-life narrative. End with 5 marketing messages that would resonate with this persona.
Brand Voice
Brand Voice Guidelines
Develop comprehensive brand voice guidelines for [BRAND]. Include: voice attributes (3-5 adjectives), tone variations by context (social media, customer service, formal communications), words to use vs. avoid, example phrases, and do's/don'ts. Provide before/after examples of copy transformation.
Social Media
Monthly Social Media Calendar
Create a 30-day social media content calendar for [BRAND] on [PLATFORMS]. Include: post themes tied to our goals of [GOALS], optimal posting times, content mix (educational, promotional, engagement, UGC), hashtag strategies, and 2-3 specific post ideas per day with captions. Consider [UPCOMING EVENTS/HOLIDAYS].
Marketing ROI
Marketing Budget Allocation Framework
Act as a CMO. Given a marketing budget of [AMOUNT] and goals of [GOALS], create a channel allocation framework. Include: recommended spend per channel, expected ROI ranges, testing budget allocation, seasonality adjustments, and KPIs to track. Consider our current performance baselines of [BASELINES].
Content Strategy
Content Pillar Strategy
Develop a content pillar strategy for [BRAND] in [INDUSTRY]. Create 5-7 content pillars that align with our expertise and audience interests. For each pillar, provide: description, example topics (10 each), content formats, SEO keyword clusters, and how it supports business goals. Include a content repurposing framework.
A/B Testing
A/B Test Hypothesis Generator
Generate 10 high-impact A/B test hypotheses for [WEBPAGE/CAMPAIGN]. For each test, provide: hypothesis statement (If we change X, then Y will happen because Z), control vs. variation description, primary metric, expected impact range, and sample size recommendation. Prioritize by potential impact and ease of implementation.
Copywriting Prompts
Headlines, ad copy, landing pages, and persuasive writing prompts.
Headlines
Headline Variations Generator
Generate 20 headline variations for [PRODUCT/TOPIC] targeting [AUDIENCE]. Include: 5 benefit-driven headlines, 5 curiosity-driven headlines, 5 urgency/scarcity headlines, and 5 social proof headlines. Each headline should be under 60 characters. Explain the psychological trigger behind each.
Landing Page
High-Converting Landing Page Copy
Write complete landing page copy for [PRODUCT/SERVICE] using the AIDA framework. Include: attention-grabbing headline and subheadline, problem agitation section, solution presentation, 5 key benefits with supporting copy, social proof section, FAQ addressing top objections, and a compelling CTA. Target conversion goal: [GOAL].
Ad Copy
Facebook/Meta Ad Copy Variations
Create 10 Facebook ad copy variations for [PRODUCT]. Include: 3 short-form (under 125 chars), 4 medium (125-250 chars), 3 long-form (250-500 chars). Each should have primary text, headline, and CTA. Test different angles: pain point, transformation, social proof, curiosity, and direct offer.
Product Descriptions
Compelling Product Description
Write a product description for [PRODUCT] that converts browsers to buyers. Include: sensory language that helps readers imagine using it, unique selling points, specifications presented as benefits, ideal use cases, and a mini-story of transformation. Length: [WORD COUNT]. Tone: [TONE]. Include SEO keywords: [KEYWORDS].
Storytelling
Brand Storytelling Framework
Create a brand story for [BRAND] using the StoryBrand framework. Define: the customer as hero, their problem (external, internal, philosophical), position our brand as guide, the plan we offer, the call to action, the stakes of failure, and the transformation/success. Write a 500-word founding story that brings this to life.
Value Proposition
Value Proposition Canvas
Create a value proposition for [PRODUCT/SERVICE]. Define customer jobs (functional, social, emotional), pains (frustrations, obstacles, risks), and gains (desired outcomes). Then map our product's: pain relievers, gain creators, and features. End with 5 value proposition statement variations following the format: "We help [WHO] do [WHAT] by [HOW] so they can [BENEFIT]."
CTAs
CTA Button & Microcopy
Generate 15 CTA button variations for [ACTION/GOAL]. For each, provide: button text, supporting microcopy (anxiety reducer), and the psychological principle it leverages. Include a mix of: action-oriented, benefit-focused, urgency-based, and conversational CTAs. All should be mobile-friendly (under 25 characters for button text).
Rewriting
Copy Improvement Analysis
Analyze and improve this copy: [PASTE COPY]. Provide: specific issues identified (clarity, persuasion, structure), an improved version with tracked changes explained, readability score improvement, and 3 alternative approaches. Maintain the same word count but increase conversion potential. Keep brand voice of [TONE].
Analytics Prompts
Data analysis, reporting, insights generation, and visualization prompts.
Tactical
Monday KPI Review — Decision Memo
I will paste weekly funnel data. Output in this exact format: 1) What changed (3 bullets, numbers only). 2) Why it likely changed (max 3 hypotheses). 3) What to do this week (3 actions with owner and expected impact). 4) What data is missing to confirm. Keep under 180 words. Do not invent numbers. Paste data: [PASTE HERE]
Tactical
Turn Analysis Into Slide Bullets
Turn this analysis into 4 bullets for a slide. Each bullet: start with what happened, end with what we're doing about it. Keep under 15 words per bullet. Then rewrite the same 4 bullets for a non-technical [MARKETING/SALES/EXEC] audience. Paste analysis: [PASTE HERE]
Data Interpretation
Marketing Data Analysis
Analyze this marketing performance data: [PASTE DATA]. Provide: executive summary, key trends identified, anomalies or concerns, performance vs. benchmarks, root cause hypotheses for underperformance, top 5 actionable recommendations with expected impact, and suggested next analyses to conduct.
SQL Generation
SQL Query Generator
Write a SQL query to [DESCRIBE ANALYSIS]. The database has these tables: [LIST TABLES AND KEY COLUMNS]. Include: the query with comments explaining each part, sample output structure, edge cases handled, performance optimization notes, and 2 variations for related analyses.
Dashboard Design
Analytics Dashboard Blueprint
Design a [TYPE] dashboard for [AUDIENCE]. Include: dashboard objective, 8-12 KPIs with definitions and formulas, recommended visualizations for each metric, filter/drill-down capabilities needed, data refresh frequency, alert thresholds, and a wireframe layout description. Consider the decisions this dashboard should enable.
Report Writing
Executive Summary Generator
Write an executive summary for this data analysis: [PASTE FINDINGS]. Structure it for [C-SUITE/MANAGER] audience. Include: headline insight (1 sentence), key metrics with context, what's working, what's concerning, recommended actions with business impact, and next steps. Keep it under 300 words. Use confident, decisive language.
Trend Analysis
Trend Identification Framework
Analyze trends in this time-series data: [PASTE DATA]. Identify: overall direction (growth/decline rate), seasonality patterns, cyclical patterns, notable inflection points with potential causes, leading indicators, and forecast for next [TIME PERIOD] with confidence intervals. Recommend actions based on trend trajectory.
KPI Definition
KPI Framework Development
Create a KPI framework for [BUSINESS FUNCTION/GOAL]. Define 10-15 KPIs organized by: leading vs. lagging indicators, strategic vs. operational metrics. For each KPI include: definition, formula, data source, measurement frequency, benchmark/target, owner, and actions triggered by different performance levels.
Cohort Analysis
Cohort Analysis Design
Design a cohort analysis for [BUSINESS QUESTION]. Define: cohort segmentation criteria (time-based, behavior-based), metrics to track per cohort, analysis time frame, visualization approach, statistical significance considerations, and expected insights. Provide SQL pseudocode for the analysis structure.
Anomaly Detection
Data Anomaly Investigation
I noticed [ANOMALY] in our data. Help me investigate: create a hypothesis tree of possible causes (technical, external, behavioral), suggest data points to examine for each hypothesis, recommend validation steps, provide a template for documenting findings, and outline the communication plan if the anomaly is significant.
🔬 Data Science Prompts
Machine learning, Python coding, statistical analysis, and modeling prompts.
Python Code
Data Cleaning Pipeline
Write a Python data cleaning pipeline for a dataset with: [DESCRIBE DATA]. Include: loading data, handling missing values (with multiple strategies), outlier detection and treatment, data type conversions, feature engineering suggestions, validation checks, and logging. Use pandas best practices and include docstrings.
ML Model
ML Model Selection Guide
I need to predict [TARGET] using [FEATURES] with [DATA SIZE] records. Help me: compare 4-5 suitable algorithms with pros/cons, recommend preprocessing steps for each, suggest hyperparameter tuning approaches, define evaluation metrics and why, outline a validation strategy, and identify potential pitfalls.
Visualization
Data Visualization Code
Create Python visualization code to show [INSIGHT] from this data structure: [DESCRIBE]. Use matplotlib/seaborn. Include: publication-quality formatting, clear labels and titles, color palette for accessibility, multiple subplot layout if needed, annotations for key points, and export settings for presentations.
Statistical Test
Statistical Test Selection
I want to test if [HYPOTHESIS]. My data: [DESCRIBE DATA TYPE, SAMPLE SIZE, DISTRIBUTION]. Recommend: the appropriate statistical test with justification, assumptions to verify, Python code using scipy/statsmodels, how to interpret results, effect size calculation, and how to report findings in a business context.
Feature Engineering
Feature Engineering Ideas
Given these raw features for predicting [TARGET]: [LIST FEATURES], suggest 20+ engineered features. For each: name, calculation logic, intuition for why it might be predictive, Python implementation, and feature importance testing approach. Group by feature type (interaction, aggregation, time-based, etc.).
Model Explanation
ML Model Explanation for Stakeholders
Explain this [MODEL TYPE] model results to non-technical stakeholders. Model predicts [TARGET] with [ACCURACY]. Top features: [LIST]. Create: an executive summary, analogy-based explanation, visual description suggestions, business implications, limitations and caveats, and recommended next steps. Avoid jargon.
Email Marketing Prompts
Subject lines, email sequences, personalization, and CRM prompts.
Subject Lines
Subject Line A/B Test Generator
Generate 20 subject line variations for an email about [TOPIC] to [AUDIENCE]. Create 4 variations each of: curiosity-driven, benefit-focused, urgency/scarcity, personalization-based, and question format. Each under 50 characters. Include preview text suggestions. Rate each for spam risk (low/medium/high).
Welcome Sequence
Welcome Email Sequence
Create a 5-email welcome sequence for new [PRODUCT/SERVICE] subscribers. For each email: subject line + preview text, send timing (day/hour), goal of the email, full copy, CTA, and success metrics. Progress from welcome → value delivery → social proof → engagement → conversion. Include personalization tokens.
Re-engagement
Win-Back Campaign Sequence
Create a win-back email sequence for customers who haven't [ENGAGED/PURCHASED] in [TIME PERIOD]. Design 4 emails with: emotional hook (we miss you), value reminder, special offer, and final attempt. Include subject lines, timing, copy, and criteria for suppressing non-responders. Consider why they may have disengaged.
Personalization
Dynamic Content Personalization
Create personalized email content blocks for [EMAIL TYPE] based on these segments: [LIST SEGMENTS]. For each segment, provide: personalized headline, body copy variation, product recommendations logic, CTA variation, and image suggestions. Show the default version and 3+ segment-specific versions.
Abandoned Cart
Abandoned Cart Recovery Sequence
Design a 3-email abandoned cart sequence for [PRODUCT TYPE]. Email 1: Reminder (1 hour), Email 2: Overcome objections (24 hours), Email 3: Incentive (48 hours). For each: subject line, copy, product block suggestions, trust elements to include, and A/B test recommendations. Include SMS follow-up option.
Newsletter
Newsletter Content Plan
Create a newsletter template and 4 weeks of content for [BRAND/TOPIC]. Include: recurring sections (define 4-5), content mix ratio, optimal length, subject line formulas, send time recommendations, engagement hooks, and growth tactics. For each week, provide specific content ideas with headlines.
Product Prompts
PRDs, user stories, roadmapping, and product strategy prompts.
Tactical
Feature Request to Sprint-Ready Spec (30 Min)
Convert this request into a spec. Request: [PASTE]. Output: A) One-sentence problem. B) In-scope / out-of-scope. C) 5 user stories with Given/When/Then acceptance criteria. D) Top 5 edge cases. E) Telemetry events to track. Then create two options: Lean (2 weeks) vs Robust (6 weeks) with tradeoffs. End with: What could fail in onboarding, permissions, data accuracy, or support? Mitigation for each.
Tactical
Pre-Launch Risk Pressure Test
For feature [FEATURE] launching to [USERS], pretend you are Support, QA, and Compliance. What will break first? Output a table: Risk | Severity | Mitigation | Owner. Then define events, properties, and success threshold for the first 14 days post-launch.
PRD
Product Requirements Document
Create a PRD for [FEATURE]. Include: problem statement with data, user personas affected, jobs to be done, success metrics, requirements (must have, should have, nice to have), user flows, edge cases, technical considerations, dependencies, risks, and launch checklist. Format for engineering handoff.
User Stories
User Story Generator
Generate user stories for [FEATURE/EPIC]. Create 10-15 stories following: "As a [user type], I want [action] so that [benefit]." For each story include: acceptance criteria (3-5 specific conditions), story points estimate, dependencies, and potential edge cases. Group by user type and prioritize by value/effort.
User Research
User Interview Analysis
Analyze these user interview notes: [PASTE NOTES]. Extract: key themes (group related insights), user pain points ranked by frequency and severity, desired outcomes, surprising findings, quotes that capture sentiment, and recommended actions. Create an affinity map structure and suggest follow-up questions.
Roadmap
Quarterly Roadmap Planning
Help plan Q[X] roadmap for [PRODUCT]. Given these inputs: company goals [GOALS], current metrics [METRICS], user feedback themes [THEMES], and tech debt items [ITEMS]. Create: theme-based roadmap, prioritization rationale (RICE scores), resource allocation suggestions, dependencies map, and risk assessment. Include "not doing" list with reasoning.
Feature Prioritization
RICE Scoring Framework
Apply RICE scoring to these feature ideas: [LIST FEATURES]. For each feature, estimate: Reach (users/quarter), Impact (0.25-3x scale), Confidence (%), Effort (person-weeks). Calculate RICE scores, rank features, and provide reasoning for estimates. Identify any features that need more data before scoring.
Competitor Analysis
Product Competitive Analysis
Create a competitive analysis for [PRODUCT] vs [COMPETITORS]. Compare: feature parity matrix, pricing models, target user segments, UX strengths/weaknesses, market positioning, recent product updates, and user review sentiment. Identify: competitive moats, vulnerability areas, and differentiation opportunities.
🗺️ Customer Journey Prompts
Journey mapping, experience design, and touchpoint optimization prompts.
Journey Mapping
Customer Journey Map
Create a detailed customer journey map for [PERSONA] using [PRODUCT/SERVICE]. Map stages: Awareness → Consideration → Purchase → Onboarding → Usage → Advocacy. For each stage include: user goals, touchpoints, actions, thoughts, emotions (high/low), pain points, and opportunities for improvement. Add moments of truth.
Touchpoint Audit
Touchpoint Experience Audit
Audit customer touchpoints for [BUSINESS]. List all touchpoints across channels (digital, physical, human). For each: rate current experience (1-5), identify friction points, benchmark against expectations, suggest quick wins and longer-term improvements. Prioritize by impact on satisfaction and feasibility.
Onboarding
User Onboarding Flow Design
Design an onboarding flow for [PRODUCT] targeting [USER TYPE]. Define: the "aha moment" to reach, steps to get there (max 5), information to collect, features to highlight, gamification elements, progress indicators, and re-engagement triggers for drop-offs. Include copy for each step and success metrics.
Churn Prevention
Churn Prediction & Prevention
Create a churn prevention framework for [PRODUCT/SERVICE]. Define: early warning signals (behavioral triggers), health score components, intervention points, retention plays for each risk level, win-back strategies, and exit interview questions. Include automation rules for triggering interventions.
Database Marketing Prompts
Audience segmentation, propensity modeling, campaign targeting, and CRM analysis prompts.
Segmentation
RFM Segmentation Framework
Build an RFM segmentation model for [BUSINESS/DATABASE]. Given: customer purchase history with recency, frequency, and monetary value fields. Output: 5–7 meaningful segments with labels (e.g., Champions, At-Risk, Hibernating), thresholds for each dimension, recommended actions per segment, and expected response rates. Include SQL pseudocode for segment assignment. Best in ChatGPT or Claude.
Propensity
Propensity-to-Buy Model Brief
Create a brief for building a propensity-to-buy model for [PRODUCT/OFFER]. Include: input features to consider (demographic, behavioral, transactional), model type recommendations with rationale, validation approach, score banding strategy (deciles), business rules for campaign suppression, and a test vs. control design to measure lift. Best in Claude or ChatGPT Thinking for complex reasoning.
Tactical
Database Audit — 30-Min Workflow
Audit this customer database extract for campaign readiness. Paste data summary or column list: [PASTE]. Check: completeness of key fields (email, postal, opt-in status), duplicate logic, data currency (last updated), segment distribution, and compliance flags. Output a table: Field | Completeness % | Issue | Fix Required. Then recommend top 3 data quality actions before any campaign send.
Suppression
Suppression & Compliance Rules Builder
Build a suppression logic document for [CAMPAIGN TYPE] targeting [AUDIENCE]. Define: hard suppression rules (opt-outs, unsubscribes, legal blocks), soft suppression rules (recent buyers, complaint history, fatigue thresholds), frequency capping logic, reachable vs contactable definitions, and GDPR/CCPA considerations. Format as a decision tree for campaign ops to follow.
Lookalike
Lookalike Audience Design
Design a lookalike audience strategy for [PRODUCT/BRAND] using our best customers as seed. Step 1: Define seed audience criteria (top 20% by LTV). Step 2: Identify key characteristics for matching (demographics, behavioral signals, purchase patterns). Step 3: Recommend platforms and match rates. Step 4: Define audience size tiers vs. precision tradeoff. Step 5: Measurement plan to validate lookalike quality vs. broad targeting.
LTV
Customer Lifetime Value Analysis
Analyze customer lifetime value for [BUSINESS]. I will paste cohort or purchase data: [PASTE DATA]. Output: average LTV by segment, top 10% vs. bottom 10% profile differences, payback period by acquisition channel, LTV:CAC ratios by channel, and 3 recommendations for shifting budget toward highest LTV segments. Do not invent numbers. Flag any data gaps.
Reasoning workflows
Reasoning-Era Prompts
Prompts designed for reasoning / Thinking models and advanced chain-of-thought workflows. Use these when accuracy matters more than speed.
When to use these: These prompts are optimized for ChatGPT's Thinking mode, Claude's extended thinking, or any reasoning-capable model when doing complex analysis, logic, or multi-step tasks. For faster tasks, regular chat models (ChatGPT, Gemini) work fine. See our reasoning model guide →
Logic Check
Decision Stress-Test
I am considering this decision: [DESCRIBE DECISION]. Before giving a recommendation, work through: 1) What assumptions is this decision built on? 2) Which assumptions are most likely to be wrong? 3) What is the second-order consequence if the top assumption fails? 4) What would a confident critic say? 5) What information, if available, would change the recommendation? Then give a verdict with confidence level (High / Medium / Low) and a one-sentence rationale. Best in ChatGPT Thinking or Claude extended thinking.
Chain-of-Thought
Complex Trade-off Analysis
Analyze this trade-off: [DESCRIBE OPTIONS AND CONTEXT]. Work step by step. First: define what we are optimizing for. Second: list criteria for evaluation with weights. Third: score each option against each criterion — show your work, don't jump to totals. Fourth: identify any criteria where the scoring is highly uncertain. Fifth: give a final recommendation with a one-paragraph rationale. Do not skip steps. Best in ChatGPT Thinking or Claude extended thinking.
Multi-Step Analysis
Root Cause Deep Dive
We observed this problem: [DESCRIBE METRIC DROP OR ISSUE]. Work through a structured root cause analysis. Step 1: List possible causes grouped by category (technical, behavioral, external, data quality). Step 2: For each cause, state what evidence would confirm or rule it out. Step 3: Rank causes by probability given what we know. Step 4: Recommend the top 3 validation queries or checks. Do not guess at causes without stating what confirms each. Best in ChatGPT Thinking or Claude extended thinking.
Scenario Planning
Three-Scenario Business Planning
For [BUSINESS DECISION / INITIATIVE], build three scenarios: Base (most likely), Bull (best reasonable case), Bear (realistic downside). For each scenario: state the key assumption that drives it, estimate the financial or operational impact with ranges, name the earliest leading indicator you would see if that scenario is unfolding, and define a trigger point for each scenario where strategy should change. Be specific with numbers; note where estimates are uncertain. Best in ChatGPT Thinking or Claude extended thinking.
Code Review
Deep Code & Logic Audit
Review this code/logic for correctness: [PASTE CODE OR LOGIC]. Do not just check syntax. Work through: 1) Edge cases that could cause wrong output. 2) Off-by-one or boundary errors. 3) Assumptions in the logic that might not hold in production data. 4) Performance issues at scale. 5) Security or data exposure risks. For each issue: severity (Critical / Medium / Low), explanation, and suggested fix. Best in ChatGPT Thinking — takes longer but catches more.
Reasoning
AI Role Displacement Self-Assessment
My role is [JOB TITLE] at a [COMPANY TYPE]. My key responsibilities are: [LIST 5–7 TASKS]. Reason through: 1) Which of these tasks can current AI (2026) fully automate? 2) Which can AI significantly accelerate but still require human judgment? 3) Which remain human-only for now? 4) What new skills should I build to stay ahead of automation in 12–24 months? Be honest, not reassuring. Use current capabilities, not speculative future AI. Best in ChatGPT Thinking or Claude for nuanced judgment.
Job search
Career prompts
Templates for resumes, LinkedIn, interviews, and reframing your experience. Pair with our resume and LinkedIn guide and career hub.
Resume
Resume bullets from raw experience
I am applying for [ROLE TITLE] in [INDUSTRY]. Below is my work history in rough notes: [PASTE BULLETS OR PARAGRAPH]. Rewrite as 5–7 resume bullets. Use strong verbs, quantify impact where I gave numbers, and do not invent metrics. If a claim needs a number I did not provide, phrase it qualitatively or mark [VERIFY]. Tailor emphasis to keywords from this job description: [PASTE JD OR KEY PHRASES].
LinkedIn
LinkedIn About section
Write a LinkedIn About section for me. Current title: [TITLE]. Target audience: [e.g. hiring managers in SaaS marketing]. Tone: [professional but conversational / sharp / warm]. Include: what I do today, one concrete win, what I am looking for next, and a single line on how I use AI at work (honest, not buzzwordy). Max 2600 characters. Here is background: [PASTE SHORT BIO OR RESUME SUMMARY].
Cover letter
Cover letter from job description
Job description: [PASTE JD]. My background: [2–3 sentences or paste resume summary]. Write a cover letter under 350 words. Open with why this company or role specifically (use public facts only, no flattery). Map three of my strengths to three requirements in the JD. Close with one sentence on fit. Avoid clichés like "passionate self-starter." Sound like a human.
Interview
Interview prep for one role
Role: [TITLE] at [TYPE OF COMPANY]. Job description summary: [PASTE OR SUMMARIZE]. Based on this, list 12 likely interview questions (mix of behavioral and role-specific). For each question, give a 4–6 sentence outline of a strong answer using the STAR method where it fits. Pull themes from my background: [PASTE 5 BULLETS]. Flag any answer that would need a real metric I should look up before the interview.
Pivot
Reframe experience for an AI-adjacent role
Current role: [TITLE]. I want to move toward [TARGET ROLE, e.g. AI product analyst / marketing ops with AI / data + GenAI]. Here are my skills and projects: [LIST]. Rewrite my positioning in three parts: (1) One paragraph "elevator" story for networking. (2) Five resume-ready bullets that emphasize tooling, data, and judgment, not hype. (3) Three gaps I should close in the next 90 days with concrete actions. Be direct if the pivot is a stretch.
Negotiation
Salary conversation talking points
I received an offer (or expect one) for [ROLE] in [LOCATION OR REMOTE]. Base range I have in mind: [RANGE if known]. Benefits that matter to me: [LIST]. Help me draft: (1) A short script to ask for time to review. (2) 4 bullet points to justify asking at the top of the range without sounding entitled. (3) One paragraph if they say the budget is fixed. (4) Questions to ask about equity, bonus, and title. This is for planning only, not legal advice.
Networking
Cold outreach message
I want to message [ROLE, e.g. a product lead] at [COMPANY OR TYPE OF COMPANY] on LinkedIn. Goal: [informational chat / referral / advice on breaking in]. My background in one line: [LINE]. Draft a message under 90 words. No "I hope this finds you well." One specific reason I picked them (their post, talk, or company move). One clear ask. Humble tone.
Portfolio
GitHub README for a small project
I built [WHAT IT DOES] using [STACK OR TOOLS]. Audience: recruiters and hiring managers who are not deeply technical. Write a README.md with: project name and one-line description, problem solved, what I built (3 bullets), how to run it or view it (or "demo link: ..."), what I learned, and "Possible next steps." Keep jargon low. I will paste rough notes: [PASTE].
Story
"Tell me about yourself" in 60 seconds
Help me script a 60-second "tell me about yourself" for interviews. Current situation: [WHERE I AM NOW]. Target: [ROLE OR INDUSTRY]. Strengths I want highlighted: [3 THINGS]. One honest reason for the move. Structure: past (10 sec), present (25 sec), future (25 sec). Conversational, not memorized-sounding. I will practice aloud, so avoid tongue-twisters.
Skills
Transferable skills audit
I work as [ROLE] and want to understand what transfers to [TARGET FIELD OR ROLE]. List my likely transferable skills from this job description I wrote: [PASTE]. For each skill, rate relevance High/Medium/Low for the target, and give one sentence I could use on LinkedIn or in an interview. Then list three skills that are real gaps and the cheapest way to demonstrate progress on each (course, small project, volunteer task).
🎨 General Prompts
Versatile prompts for everyday professional use.
Meeting Prep
Meeting Preparation Brief
Help me prepare for a meeting about [TOPIC] with [ATTENDEES]. Create: agenda suggestion, key talking points, potential questions I'll face with suggested answers, data/examples I should bring, decisions to drive, and follow-up items to propose. Meeting goal: [GOAL].
Presentation
Presentation Outline Generator
Create a presentation outline for [TOPIC] targeting [AUDIENCE]. Include: hook opening (story/statistic), problem framing, solution/recommendation, supporting evidence (3 pillars), objection handling, call to action, and closing. Suggest visuals for each slide. Time: [DURATION]. Goal: [ACTION YOU WANT].
Problem Solving
Problem-Solving Framework
Help me solve this problem: [DESCRIBE PROBLEM]. Walk me through: problem definition (5 Whys), stakeholder map, constraint identification, solution brainstorm (10+ ideas), evaluation criteria, top 3 solutions with pros/cons, recommended approach, implementation plan, and success metrics.
Communication
Difficult Conversation Prep
Help me prepare for a difficult conversation about [SITUATION] with [PERSON/ROLE]. Provide: opening script that's direct but empathetic, key points to make, anticipated responses and how to handle each, de-escalation phrases, desired outcome, and follow-up steps. Tone should be professional and constructive.
Put These Prompts to Work
Now that you have powerful prompts, level up your AI skills with our structured courses and certifications.