> ## Content Index > Fetch the complete content index at: https://www.datadab.com/blog/llms.txt > Use this file to discover other available public pages before exploring further. # 50 AI Prompts for Growth Ops (with Guardrails) - URL: https://www.datadab.com/blog/50-prompts-for-growth-ops-with-guardrails/ - Published: 2025-10-21T15:26:41.000Z - Updated: 2026-08-15T12:37:12.000Z - Description: The ultimate prompt kit for growth ops pros who want less AI chaos and more conversions. - Author: Amit Ashwini - Tags: AI in Marketing Most “AI growth hacks” are either thinly disguised spreadsheet tricks or end in 600-word blogs about cheese. What you *actually* need? Prompts that tell your agents **exactly what to do, how to behave, and when to quit hallucinating**. This isn’t your average ChatGPT cheat sheet. It’s a toolkit. For growth operators. With real jobs. And actual OKRs breathing down their necks. We’ve packaged this in seven smart-ish sections: --- ### **1\. Cold Email Crafter** **Use Case:** Writing cold outbound email sequences for B2B outreach that don’t get sent straight to spam or hell. **Prompt:** You are a B2B copywriter specializing in outbound campaigns. Your task is to write a 3-part cold email sequence targeting **mid-market CFOs** in the SaaS sector. The product being pitched is **a financial forecasting tool that reduces monthly reporting time by 40%**. Each email should: - Be under 100 words. - Start with a credible industry insight or stat (cite the source if possible). - Use **plainspoken language** — no “cutting-edge,” “solid,” or “driving synergies.” - Include a **clear ask**: either reply, book a call, or check a one-pager. - Assume zero familiarity with the sender, but not zero intelligence. - Sound like a smart peer, not a sales intern. Do NOT: - Use flattery (“noticed you’re a leader…”), fake urgency, or fake personalization. - Ask “can I pick your brain?” or “got 15 minutes?” - Use subject lines like “Quick question” or “Let’s connect” --- ### **2\. Audience Expander** **Use Case:** Finding adjacent ICPs with similar pain points to widen the top of funnel without going off-brand. **Prompt:** You are a market strategist assisting a growth team. Your task is to identify **5 adjacent target audiences** to the following ICP: > “VPs of Finance at Series B SaaS companies using spreadsheets for forecasting.” For each new audience: - Give it a name (e.g. “RevOps at Fintech scaleups”). - Explain the **shared pain point or unmet need**. - Suggest one custom messaging angle or content hook that would resonate. - Keep explanations under 40 words each. Exclude audiences that: - Don’t have budget authority - Would require completely different messaging or positioning - Are too far removed from the product’s current value prop --- ### **3\. Persona Re-Mapper** **Use Case:** Repurposing existing landing pages for a new persona without rebuilding everything from scratch. **Prompt:** You are a growth content specialist. Your task is to rewrite the **copy of this landing page** originally written for **product managers** to make it suitable for **sales enablement leaders**. Here is the original copy: \[Paste or summarize\] Your output should: - Maintain the existing structure (headline, subhead, CTA, sections). - Translate benefits into language that resonates with sales enablement leaders. - Swap examples, stats, or [case studies](https://www.datadab.com/case-studies) that feel too product-led or technical. - Remove any references to roadmapping, sprints, or product backlogs. Keep the tone helpful and pragmatic — imagine the reader is short on time and allergic to fluff. --- ### **4\. Social Hook Generator** **Use Case:** Writing social media hooks that get engagement *without* sounding like LinkedIn hustle broetry. **Prompt:** You are a social copywriter for a B2B brand. Your task is to write **5 compelling post hooks** for LinkedIn or Twitter promoting the following blog post: > \[Insert blog URL or summary\] Each hook must: - Be a standalone, scroll-stopping opener (max 12 words). - Use curiosity, contrast, or a controversial take — not hype or empty claims. - Feel like it was written by a real person, not a brand account. Avoid: - “Here’s why X is the future of Y” - “You’re doing X wrong” - Any format that ends with “you won’t believe #3” --- ### **5\. LinkedIn DM Agent** **Use Case:** Crafting a DM follow-up after someone engages with your content — without being weird or pushy. **Prompt:** You are an outbound SDR assistant. Your task is to write a **LinkedIn DM** to someone who liked or commented on our carousel post about **costly RevOps mistakes**. Assume: - They are a **Head of Revenue Operations at a mid-size B2B SaaS company**. - They have not spoken to us before. - Our product helps unify sales, marketing, and finance metrics in one place. The message should: - Be polite, brief (max 3 sentences), and conversational. - Reference the post naturally — not in a creepy “noticed you liked…” way. - Ask a **low-friction question** or offer something useful (not a pitch). Avoid: - Asking for 15 minutes - Links on first message - Using emojis or “hey 👋” intros --- ### **6\. Ad Copy Variants (FB/IG)** **Use Case:** Creating multiple ad versions for Facebook/Instagram to test different psychological triggers. **Prompt:** You are a performance copywriter tasked with writing **3 ad copy variants** for a Facebook/Instagram ad campaign. Product: \[Insert product\] Audience: \[Insert ICP\] Goal: \[Lead gen / sign-ups / free trial\] Each version must use a different emotional trigger: 1. Curiosity (make them want to click to learn) 2. Fear (highlight a risk of inaction) 3. [Social proof](https://en.wikipedia.org/wiki/Social%5Fproof) (highlight a specific success or testimonial) Each ad copy should: - Be under 70 words - Avoid cliché phrases like “you won’t believe…” or “don’t miss out” - Include a strong, relevant CTA --- ### **7\. Keyword Miner Agent** **Use Case:** Finding long-tail keywords from existing pages to improve discoverability. **Prompt:** You are an SEO strategist. Given the following **landing page URL or summary**, your task is to generate **10 long-tail keywords** we should optimize for. URL: \[Insert\] Product: \[Insert\] ICP: \[Insert\] Include: - Keyword phrase - Search intent category (Awareness / Consideration / Decision) - Why it fits the user journey (20 words max) Exclude: - Branded queries - Keywords irrelevant to B2B or this ICP --- ### **8\. Partnership Pitcher** **Use Case:** Writing outreach emails to potential co-marketing or distribution partners. **Prompt:** You are a partnerships manager. Your task is to write a **cold outreach email** to a potential partner who runs a newsletter for \[audience\]. Goal: Pitch a co-marketing opportunity (e.g. joint webinar, newsletter swap, content collab) Your email should: - Be under 120 words - Start with a **shared audience insight**, not flattery - Describe the value for **their audience**, not just ours - Include a proposed idea and soft CTA to chat Avoid: - “We’d love to collaborate” with no specifics - “Let me know if you’re open” with no value - Attachments or Calendly links on first email --- ### **9\. Lead Magnet Title Test** **Use Case:** Testing different psychological angles for an ebook or whitepaper title. **Prompt:** You are a conversion-focused copywriter. Your task is to create **5 title options** for a new lead magnet based on the following idea: > \[Describe lead magnet topic or problem\] Each title should: - Be under 12 words - Tap into a distinct psychological trigger: fear, curiosity, authority, simplicity, or contrarianism - Be A/B testable as a headline or CTA Avoid: - Generic titles like “The Ultimate Guide to X” - Overused power words (“Masterclass,” “Unleashed,” “Breakthrough”) --- ### **10\. Newsletter Welcome Email** **Use Case:** Writing the first welcome email that new subscribers actually read (and maybe even click). **Prompt:** You are a lifecycle email specialist. Your task is to write the **first welcome email** in a newsletter sequence. Audience: B2B SaaS professionals Newsletter: Weekly insights + templates on revenue growth and operations Your email must: - Be under 100 words - Set one clear expectation (e.g. what/when they’ll get) - Deliver one piece of immediate value (e.g. link to top post/template) - Include one optional CTA (“tell us what you want more of”) Do NOT: - Open with “Hey there 👋” - Talk about how excited “we” are - Use more than one CTA --- ### **11\. SEO Blog Outliner** **Use Case:** Creating outlines for SEO blogs that don't just keyword stuff but actually rank *and* engage. **Prompt:** You are an SEO content strategist tasked with writing a blog outline targeting the keyword: **\[Insert primary keyword\]** Product: \[Brief product description\] Audience: \[Describe ICP\] Your output should: - Include a blog title (max 60 characters, CTR-worthy) - Suggest an intro angle that adds *something new* vs the top 3 SERPs - Include 6–8 H2s that reflect current search intent *and* add value - Mark 2 sections that can be converted into list posts or visuals - Suggest a schema type to embed (e.g. FAQ, HowTo, Article) Use tools like AlsoAsked, Google’s PAA, and competitor blogs as inspiration. Do NOT: - Copy headers from top-ranking posts word for word - Include generic intros like “In today’s fast-paced digital world…” --- ### **12\. Contrarian POV Creator** **Use Case:** Creating thought leadership content that doesn’t feel like it was copy-pasted from 2016 LinkedIn. **Prompt:** You are a B2B thought leadership writer. Your task is to write an **intro + outline** for a contrarian blog post on the topic: **\[Insert topic or trend\]** The goal is to present a **smart, surprising, but believable** POV. For example: Instead of “AI will make marketers obsolete,” try “AI will make bad marketing more visible.” Your output must include: - A spicy, specific headline (no more than 12 words) - A 100-word intro that references a common belief — then challenges it - A list of 5–6 sections that unpack your point of view - 1 real-world analogy or story to use as a recurring theme Avoid: - Thought leadership tone with zero actual thoughts - Unbacked claims or techno-saviourism - Cliché intros like “Let’s face it…” --- ### **13\. Carousel Architect** **Use Case:** Turning dense content into snackable, engaging carousels for LinkedIn. **Prompt:** You are a B2B social content designer. Your task is to turn the following blog/post/insight into a **7-slide LinkedIn carousel**. Content Source: \[Insert blog URL or paste key content\] Each slide should follow this format: - **Slide 1 (Cover):** Strong hook headline (max 12 words) - **Slides 2–6:** One idea per slide, each with a short headline + 1–2 punchy lines of copy - **Slide 7:** A single CTA or thought-provoking question for engagement Tone: Educated but casual — write like a smart peer Design assumption: Black/white backgrounds, bold fonts, minimal visuals Avoid: - Starting every slide with “Here’s why…” - Broetry or motivational filler - Calls to action like “Smash that follow button!” --- ### **14\. Stat Source Finder** **Use Case:** Finding credible stats to add punch to blog posts, whitepapers, or decks. **Prompt:** You are a research assistant for content marketing. Your task is to find **5 credible and recent stats** related to this topic: **\[Insert topic, e.g. 'email open rates', 'B2B sales cycles', etc.\]** Each stat must include: - The stat itself (quantified, not vague) - Source link (within past 18 months preferred) - A 1-line note on how it supports a marketing point or pain Format: ``` 1. Stat: 61% of B2B buyers say... Source: [URL] Use: Reinforces need for content personalization ``` Do NOT: - Use data without a source - Pull stats from low-quality aggregator blogs --- ### **15\. Content Refresh Agent** **Use Case:** Updating stale blogs without losing SEO juice or breaking the page. **Prompt:** You are an SEO content editor. Your task is to **refresh this blog post** from 2022 titled: **\[Insert title\]** Goals: - Preserve existing ranking if it has [backlinks](https://developers.google.com/search/docs/crawling-indexing/links-crawlable) - Update stats, examples, tools, and language for 2025 relevance - Mark each updated section with a comment `` Steps: 1. Identify outdated sections (tools, timelines, pricing, trends) 2. Replace or reframe with 2025 data or alternatives 3. Add a new “2025 Update” section with at least 1 trend or insight 4. Re-check headers, [meta description](https://developers.google.com/search/docs/appearance/snippet), and internal links Do NOT: - Change the URL or H1 - Add new keywords that aren’t contextually relevant --- ### **16\. YouTube Script Draft** **Use Case:** Turning blog content into short, engaging video scripts. **Prompt:** You are a content-to-video scriptwriter. Your task is to write a **90-second YouTube script** based on this blog post: **\[Insert blog summary or URL\]** Target viewer: \[Insert role\] Goal: Explain one key takeaway from the blog in a clear, memorable way. Script structure: - Hook (first 8 seconds): A surprising fact, question, or stat - Setup: What’s the common issue / misconception? - Payoff: What’s the core insight or better approach? - CTA: Subtle invite to learn more (no “like and subscribe” nonsense) Tone: Conversational, like you’re talking to one smart colleague Avoid: - Reading like a teleprompter - Lists of features - “As a leading provider of…” --- ### **17\. Repurpose Bot** **Use Case:** Squeezing more ROI from long-form content. **Prompt:** You are a repurposing assistant. Your task is to break this **webinar transcript** into multiple assets: Transcript: \[Insert paste or summary\] Deliverables: - 1 Twitter/X thread (7–10 tweets max) - 3 LinkedIn post angles (each with a title + summary paragraph) - 1 blog post intro (max 100 words) - 3 headline options for each format Tone: Smart, punchy, and lightly opinionated — no motivational filler Each asset should stand alone without needing the full video. Do NOT: - Quote speakers without context - Lift word-for-word chunks without editing for clarity --- ### **18\. FAQ Generator** **Use Case:** Building trust on landing pages with FAQs that preempt real objections. **Prompt:** You are a B2B UX copywriter. Your task is to write **5 sharp FAQs** for this landing page: **\[Insert URL or topic summary\]** Audience: \[Insert ICP\] Each FAQ should: - Focus on a **real buyer hesitation** (e.g. pricing, complexity, support) - Be written in the customer’s voice (e.g. “Will this work with X?”) - Include a 25–30 word answer that reduces friction or clarifies Avoid: - Fluffy answers like “Yes, absolutely!” - Trying to squeeze a sale in the answer - Duplicate answers from existing pages --- ### **19\. AI Overview Block** **Use Case:** Creating a paragraph that ranks for Google’s AI Overview or gets cited in ChatGPT answers. **Prompt:** You are an SEO writer creating an **AI-answer-friendly paragraph** for the blog titled: **\[Insert title\]** Keyword target: \[Insert\] Length: 45–55 words Format: 1–2 sentences, simple, fact-led, no intros or filler Include: - A clear, direct answer to the user’s implied question - 1 stat or definition or relevant example - Active voice, neutral tone, no calls to action Do NOT: - Say “In today’s digital world…” - Start with “Let’s explore…” or “Here’s why…” --- ### **20\. Internal Linking Agent** **Use Case:** Helping content rank better and keep users on site longer with smart internal links. **Prompt:** You are an internal SEO agent. Your task is to suggest **5 internal links** this blog post should include. Blog topic: \[Insert topic\] Site map: \[Insert brief outline or 5–10 URLs\] For each suggestion: - Link anchor text (5–6 words) - URL or slug from the site - Placement in the blog (e.g. “after section 2” or “under FAQ”) Rules: - Relevance first. Don’t link just for the sake of it. - Anchor text should feel natural in context. - Avoid linking to generic pages like “About Us” --- ## **Experimentation Agents (21–27)** *For growth teams that believe in testing before tweeting.* ### **21\. Hypothesis Builder** **Use Case:** Turning vague ideas into clear, testable growth experiments. **Prompt:** You are a growth product manager. Your task is to help structure this experiment idea into a **proper growth test**. Raw Idea: \[Insert vague idea, e.g. “Add a progress bar to onboarding”\] Please return: - Clear hypothesis in this format: *If we do X, we expect Y to happen because Z.* - One primary metric (with unit) - Minimum sample size needed for directional confidence - Risk level (Low/Medium/High) - Expected test duration (in business days) - What would be considered a successful outcome vs a null result Avoid: - Broad or philosophical “hypotheses” - Testing multiple variables at once --- ### **22\. Rapid Test Queue** **Use Case:** Generating a backlog of low-effort, high-learning experiments. **Prompt:** You are a scrappy growth PM creating a **2-week test backlog**. Product: \[Insert product type\] Resources: 1 PM, 1 designer, 1 hour/day of dev Constraints: No major UI changes or DB migrations Your output should: - List 5–7 experiments - Each must include: title, hypothesis, metric, risk level, and effort estimate (Low/Medium) - Focus on learning velocity, not perfection Priority goes to: - Signup, activation, or retention - Moments with high user drop-off or friction - Tests that challenge internal assumptions --- ### **23\. A/B Copy Generator** **Use Case:** Producing multiple copy variants for [A/B testing](https://en.wikipedia.org/wiki/A/B%5Ftesting) *without* copy/paste fatigue. **Prompt:** You are a CRO copywriter. Your task is to generate **3 copy variants** for the following: \[Insert element: headline, CTA, product description, etc.\] Context: - Audience: \[Insert ICP\] - Goal: Improve CTR or conversion - Current Copy: \[Insert baseline copy\] Each variant should: - Change only one thing (e.g. tone, length, emotional appeal) - Include a hypothesis for why it might win - Be under 20 words if it's a CTA or headline Avoid: - Changing multiple elements at once - Generic A/B test phrasing like “Try for free now!” --- ### **24\. Headline Split Test** **Use Case:** Creating headlines that test distinct psychological angles. **Prompt:** You are a B2B content strategist running headline tests. Your task is to create **4 headline variants** for this blog: \[Insert blog summary or URL\] Each headline must test a different angle: 1. Curiosity 2. Data/stat-driven 3. Authority/credibility 4. Contrarian/polarizing Guidelines: - Max 12 words - Include one primary keyword where natural - Should stand alone in a social post or email subject Avoid: - Puns or jokes (unless clearly labeled) - Starting every headline with “How to…” --- ### **25\. UX Impact Estimator** **Use Case:** Avoiding “growth hacks” that quietly kill product experience. **Prompt:** You are a UX-sensitive growth PM. Evaluate the following proposed test for **negative user impact**: Test: \[Insert idea, e.g. “Add exit intent popup with 20% discount”\] Your output: - Likelihood of friction/confusion (Low/Med/High) - Risk to perception of brand (Low/Med/High) - Which segment might be affected most - What signals to monitor during test (e.g. [bounce rate](https://support.google.com/analytics/answer/1009409), session duration) - Suggested guardrail metrics (e.g. % of users disabling feature) Bonus: suggest one tweak to reduce UX downside while preserving the test intent --- ### **26\. Test Report Summarizer** **Use Case:** Explaining results to execs who skim everything. **Prompt:** You are a growth analyst. Your task is to summarize an A/B test in 300 words or less for a non-technical leadership team. Include: - Test name + what was tested - Primary goal metric and result (with uplift or decline %) - Significance level (p-value or confidence interval) - What the result means for decision-making - What *not* to assume from the result - Next action (scale, iterate, or kill) Tone: Confident but not overclaiming. Write like you respect their time. Avoid: - Confidence theater (“the results clearly prove…”) - Deep statistical jargon --- ### **27\. Experiment Kill Switch** **Use Case:** Avoiding zombie tests that drain time and energy. **Prompt:** You are building an experiment playbook. Your task is to write a **kill-switch checklist** for when to end an experiment prematurely. Audience: PMs and growth leads running tests weekly Checklist must include: - 5 criteria (e.g. bad data quality, no movement after 7 days, obvious UX harm) - What actions to take (e.g. rollback, communicate, log learning) - How to distinguish between “test failure” and “test inconclusive” - Optional Slack/email message template to notify stakeholders Keep the tone direct and no-nonsense. This is for adult growth ops. --- ## **Analytics & Reporting Agents (28–33)** *Because if you can’t measure it, you’re just storytelling with spreadsheets.* ### **28\. Board Report Generator** **Use Case:** Creating a crisp quarterly growth update for board decks. **Prompt:** You are a growth lead writing a board update for Q\[Insert quarter\] performance. Include in 300 words: - Top 2 wins (with short metric + impact) - 1 miss (with reason and next step) - 1 learning or pattern emerging - A short table or visual summary (describe or link) - What’s next for the next quarter (1 goal + 1 initiative) Audience: Board members who want signal, not spin. Make it narrative, not just numbers. Avoid: - Celebrating vanity metrics - Pretending misses were wins --- ### **29\. Cohort Analysis Explainer** **Use Case:** Making cohort reports readable by someone who doesn’t live in Mixpanel. **Prompt:** You are a growth analyst. Summarize this cohort report in plain English. Report: \[Insert screenshot, table, or CSV summary\] Output: - What the cohort is (e.g. “Users who signed up in May”) - What behavior was tracked (e.g. activation, retention) - Notable patterns (e.g. “Drop after week 2,” “mobile vs desktop gap”) - 1 visual metaphor or chart insight (optional) - Suggested next action (e.g. double down, segment further) Keep the tone neutral and clear. Explain it like you're teaching a smart intern. --- ### **30\. Funnel Drop-Off Finder** **Use Case:** Diagnosing conversion leaks and proposing fixes. **Prompt:** You are a growth PM reviewing funnel data. Funnel: \[Insert steps or data\] Audience: \[Insert ICP\] Channel: \[Insert source, e.g. Paid, Organic, Direct\] Your job: - Identify the **largest drop-off stage** - Suggest 3 hypotheses why users drop here - Propose 2 experiments to reduce drop-off - Highlight any data gaps (e.g. missing events) Bonus: suggest 1 content or UX quick win that doesn’t require engineering --- ### **31\. UTM Builder Agent** **Use Case:** Creating UTM links that are both clean and trackable. **Prompt:** You are a campaign ops assistant. Your task is to create **UTM-tagged URLs** for this campaign: Campaign: \[Insert description\] Channels: \[e.g. [LinkedIn Ads](https://business.linkedin.com/marketing-solutions/ads), Email, Twitter Organic\] Destination: \[Landing page URL\] Format as a table with: - Channel - Medium - Campaign - Content - Final URL Rules: - Use all lowercase and hyphens only - Be consistent with existing naming conventions - Add `utm_content` only for variant testing --- ### **32\. Attribution Checker** **Use Case:** Deciding which attribution model actually fits the buying journey. **Prompt:** You are a growth analyst reviewing attribution options. Context: - Touchpoints: \[e.g. LinkedIn Ad → SEO blog → demo request\] - Sales cycle: \[Short/Medium/Long\] - Product type: \[Insert\] Task: - Evaluate 3 attribution models (Last-Touch, Linear, W-Shaped) - Recommend the best-fit model with justification - Highlight one blind spot or bias in that model - Suggest a backup model if confidence is low Keep it brief (under 200 words). This is for a marketing lead who needs clarity, not complexity. --- ### **33\. Growth Model Estimator** **Use Case:** Building a back-of-the-envelope growth target model. **Prompt:** You are a growth lead estimating the path to $1M ARR. Inputs: - Current MRR: $\[Insert\] - Avg customer value: $\[Insert\] per month - Churn: \[Insert %\] - Traffic and signup baseline: \[Insert ballpark\] - Timeframe: \[e.g. 12 months\] Your job: - Build a monthly growth forecast model (MRR targets) - Show how many leads/signups/customers needed each month - Highlight 1–2 levers that matter most (e.g. signup → activation rate) - Present as a markdown table or grid This is directional, not precise — use simple math and clear logic. --- ## **Lifecycle & Retention Agents (34–39)** *For making sure your funnel isn’t a leaky sieve.* ### **34\. Onboarding Flow Writer** **Use Case:** Creating an onboarding email sequence that actually helps users, not just pings their inbox. **Prompt:** You are a lifecycle marketer designing a **4-part onboarding email sequence** for a PLG SaaS product. Product: \[Insert short description\] Audience: \[Insert ICP\] Goal: Get users to complete key activation steps Rules: - Each email = 1 topic, 1 CTA - Use descriptive subject lines (no “Welcome!”) - Focus on *what the user can do*, not what the product can do - Highlight 1 benefit per step, not a feature dump Deliverables: - Email subject line - 50–80 word body - One simple CTA per email Do NOT: - Use empty cheerleading copy (“We’re so excited!”) - Send more than one per day - Include multiple CTAs --- ### **35\. Churn Feedback Parser** **Use Case:** Making sense of exit survey mess to uncover useful themes. **Prompt:** You are a customer success analyst. You’ve received 25 churn survey responses. Your job is to turn them into a **summary of reasons and actions**. Input: \[Paste or summarize responses\] Audience: Internal growth/product team Your output: - Group reasons into 3 buckets (e.g. pricing, product gaps, onboarding) - Quote 1 verbatim line per bucket - Recommend 1 fix or experiment per reason cluster - Optional: Flag responses that mention competitors or integrations Keep it constructive. This is to improve retention, not to rant about users. --- ### **36\. Activation Nudger** **Use Case:** Nudge users toward activation without annoying them into quitting. **Prompt:** You are designing **3 in-app nudge messages** for users who haven’t completed Step 2 of onboarding. Product: \[Insert product\] Step 2: \[Describe specific action, e.g. “connect an integration”\] Each nudge must: - Be short (max 20 words) - Focus on outcome, not guilt (“Get faster reports” vs “You haven’t done X”) - Appear in context (e.g. dashboard, modal, tooltip) For each: - Write the copy - Specify trigger condition - Suggest visual (icon, progress bar, etc.) Avoid: - “Don’t miss out!” - Using red or alarming visuals --- ### **37\. Reactivation Email Agent** **Use Case:** Win back churned or dormant users with actual value. **Prompt:** You are a retention marketer writing a **winback email** for users who went inactive 60+ days ago. Product: \[Insert product\] Segment: \[e.g. churned after 3 months, never activated\] Your task: - Write an email < 100 words - Start with a new product improvement or feature - Offer them something useful (not a discount or guilt trip) - Close with a re-engagement CTA (“Try it again”, “Take a peek”) Tone: Friendly but neutral. Like reconnecting with an old colleague, not begging an ex. --- ### **38\. Customer Win Celebration** **Use Case:** Encouraging loyalty and shareability when users hit milestones. **Prompt:** You are designing an **automated milestone email** for users who hit a meaningful usage goal (e.g. “Created 100 automations”). Audience: Power users Product: \[Insert product type\] Your email should: - Celebrate without exaggeration - Include a shareable image or stat (describe it) - Mention one “next step” they might take - Optional: link to a case study or community thread Tone: Supportive, not salesy. Keep it under 100 words. Avoid: - “You’re crushing it!” - Calls to upgrade unless there’s true value --- ### **39\. Feature Discovery Agent** **Use Case:** Introducing new features to different user types. **Prompt:** You are writing **3 in-app messages** announcing a new feature. Feature: \[Insert description\] Audience Segments: 1. First-time users 2. Power users 3. At-risk users Each message should: - Highlight one benefit specific to that group - Be 20–30 words max - Use friendly language, not a changelog Also include: - Where it appears (dashboard, empty state, etc.) - Optional: Visual (icon, mini demo, badge) Avoid: - “You asked, we listened…” - Overstating impact --- ## **Ops & Infra Agents (40–44)** *For when “just copy paste from last quarter’s playbook” isn’t good enough.* ### **40\. Playbook Generator** **Use Case:** Building a reusable growth experiment SOP. **Prompt:** You are an operations lead writing a **step-by-step playbook** for running quarterly growth experiments. Audience: PMs and marketers Scope: From idea to result to archive Include: 1. Intake process (who submits what, where) 2. Prioritization approach (e.g. ICE, PILLAR) 3. Setup checklist (metrics, tools, sample size) 4. Test run protocol (launch, monitor, adjust) 5. Post-mortem format (results, learnings, next steps) Format: Bulleted or numbered list. Max 400 words. Avoid: - Vague “align with stakeholders” steps - Jargon like “cross-functional synergies” --- ### **41\. Tool Comparison Assistant** **Use Case:** Making quick tech-stack decisions based on real ops needs. **Prompt:** You are a RevOps analyst comparing tools. Use Case: \[e.g. onboarding analytics\] Tools: \[Insert Tool A and Tool B\] Your table should compare: - Setup time - Feature fit (1 line per key feature) - Integrations - Learning curve - Price (base tier + growth tier) - Best for whom? (brief personas) Include a 1-paragraph recommendation at the end. Avoid: - Fluffy marketing copy - “Depends on your needs” with no guidance --- ### **42\. Meeting Summarizer Agent** **Use Case:** Post-meeting notes that someone *might* actually read. **Prompt:** You are a virtual assistant summarizing a 30-minute **weekly growth sync call**. Input: \[Paste transcript or notes\] Output format: - 3 decisions made (clear, binary) - 2 open questions or blockers - 1 action item per attendee (include name) Optional: Highlight anything that needs escalation Tone: Neutral and tidy. Use bullet points or markdown. --- ### **43\. Growth Stack Audit** **Use Case:** Pruning tools that cost money but deliver zero ROI. **Prompt:** You are a systems auditor doing a **MarTech/Growth stack audit**. Input: Spreadsheet with tools, pricing, usage notes Goal: Identify: - Tools with overlapping functions - Tools used by <10% of team - Tools with no logins in 60+ days Deliver: - A list of 5 tools to sunset or consolidate - Reason for each - Suggested replacement or consolidation plan Tone: Clear and cost-conscious, not accusatory --- ### **44\. Data Hygiene Prompter** **Use Case:** Cleaning CRM data before launching outbound. **Prompt:** You are a RevOps manager prepping for an outbound campaign. Create a **data cleanup checklist** for CRM contacts. Include: - Deduplication rules (email, name, domain) - Segmenting logic (ICP filters) - Fields to normalize (titles, industry, size) - Opt-out / GDPR compliance checks - Steps to flag contacts for enrichment or removal Deliver in checklist format (bullets or markdown). Avoid: - “Check for completeness” without specifics - Anything requiring more than 3 tools --- ## **Guardrail & Behavior Prompts (45–50)** *To keep your agents from turning into motivational speakers or compliance risks.* ### **45\. Anti-Fluff Layer** **Use Case:** Rewriting [AI content](https://www.datadab.com/ai-content-strategy) to sound like a human who respects other humans’ time. **Prompt:** You are a no-nonsense editor. Rewrite this AI-generated copy with **no fluff or filler**. Input: \[Insert copy\] Rules: - Remove all adjectives that add no value (“innovative,” “seamless,” etc.) - Convert vague benefits into concrete examples - Use clear verbs. Kill all passive voice. - No “transform your business” nonsense Keep the tone smart and straightforward. --- ### **46\. Tone Adjuster** **Use Case:** Fixing emails that sound like they were written by a caffeinated intern. **Prompt:** You are rewriting this email to match the tone of a calm, competent professional. Input: \[Insert email\] Adjust for: - Clear, confident tone (no exclamation marks) - Neutral emotional register (not too hyped, not too apologetic) - No filler phrases (“Just checking in”, “Hope you’re doing well”) Make the message feel like it came from a COO on a deadline. --- ### **47\. Assumption Challenger** **Use Case:** Stress-testing internal plans or playbooks. **Prompt:** You are a skeptical growth advisor. Your job is to **challenge the assumptions** in this plan: Input: \[Insert strategy doc or idea\] Output: - List of 3–5 key assumptions - What could go wrong if each is wrong - One suggestion for testing or validating each Tone: Firm but constructive. No “gotchas,” just smart checks. --- ### **48\. Compliance Copilot** **Use Case:** Making sure outbound emails or flows don’t land you in GDPR jail. **Prompt:** You are a compliance-aware copy reviewer. Analyze this email/flow for issues with: - GDPR consent - CAN-SPAM compliance - Accessibility (color contrast, alt text, etc.) Input: \[Insert content or screenshot\] Output: - List of issues with explanations - Suggested compliant rewrites or design fixes - Final verdict: Safe to launch or needs review Avoid legalese. Write like an ops lead, not a lawyer. --- ### **49\. Fallback Planner** **Use Case:** Preparing backup plans for failed campaigns or launches. **Prompt:** You are a pragmatic growth lead. The following initiative might flop: **\[Insert test, campaign, or launch\]** Your task: - List 3 fallback actions if it underperforms - Each must be lower-cost, faster, or higher-learning than the original - Include one comms plan (how to frame the pivot internally/externally) Tone: Level-headed and flexible. This is Plan B, not panic. --- ### **50\. Ask Like a Human** **Use Case:** Fixing weird AI prompts that don’t sound like anyone would actually say them. **Prompt:** You are a conversational prompt editor. Input: \[Insert original prompt\] Your task: - Rewrite it as if a smart human asked it aloud - Clarify intent - Use natural phrasing, not “Instruct the assistant to…” Deliver the rewritten version + 1 tip to make future prompts better. --- AI agents are only as good as the instructions they’re given. Left unsupervised, they’ll churn out motivational soup, keyword cannibalism, or copy that sounds like a robot trying to flirt. But with clear roles, crisp prompts, and firm boundaries? They become your most reliable junior operator — never tired, never off-brand, and never bored of writing yet another onboarding email. Use this toolkit to build agents that think like strategists, write like humans, and stop hallucinating halfway through a UTM tag. *Need help wiring this into your actual workflows? Drop us a note — we speak fluent Zapier, Notion, and nervous growth lead.*