Most "AI prompts for marketing" lists are useless. They give you a one-line prompt like "Write a Facebook ad for my product" and call it a day. The output is generic. The output sounds like every other AI-written ad on the internet. The output does not convert.
This is not that list. Every prompt below is one I have actually used in client work or my own marketing - for SEO content, paid ads, email sequences, and audience research. Each one follows a structure that forces AI to give you a specific, usable result rather than vague filler.
You can copy them verbatim. You can adapt them. You can stack them into workflows. The only thing you must not do is paste them in, accept the first output, and ship.
Why Most Marketing AI Prompts Fail
The reason most marketing prompts produce mediocre output has nothing to do with the AI model. It has to do with three missing inputs: role, context, and format.
If you ask ChatGPT "write a marketing email for my course", you have given it no role to inhabit, no context about the audience or product, and no format to follow. It will reach for the most average example of "marketing email" in its training data - which is exactly why the output sounds average.
Compare that with: "You are a senior email copywriter who has worked on B2B SaaS launches. Write a marketing email for a UK-based course teaching working professionals how to use AI tools at work. The audience is mid-career, time-poor, and sceptical of AI hype. Use a single-idea structure: hook, story, lesson, CTA. Maximum 180 words. Subject line under 45 characters."
Same task. Completely different output. The second one is usable.
The R-C-T-F Structure for Marketing Prompts
I use a four-part structure for almost every marketing prompt. Get this in your head and the rest of the article makes sense.
- Role - who the AI is pretending to be ("senior performance marketer", "B2B copywriter", "SEO strategist")
- Context - who you are, who the audience is, what you sell, what stage of awareness they are at
- Task - the specific job to be done
- Format - exact output structure (length, sections, tone, examples)
Every prompt below contains these elements. When you adapt them, replace the bracketed placeholders with real specifics. The more specific you are, the better the output.
5 Prompts for Content Calendars
1. The 30-Day Content Calendar
You are a content strategist for [INDUSTRY] brands. Generate a 30-day content calendar for [BRAND], targeting [AUDIENCE]. The brand voice is [3 ADJECTIVES]. Output a table with: date, channel (LinkedIn / blog / email / Instagram), topic, format (carousel / long-form / short / video), and the single business goal each post serves. Cluster posts into 4 weekly themes that build on each other.
2. Content Pillar Architecture
Act as a content strategist. My business is [DESCRIPTION]. My ideal customer is [PERSONA]. Propose 4 content pillars I should own across blog, social, and email for the next 12 months. For each pillar, give: the strategic reason it matters, 8 specific topic ideas, and the search intent each topic targets.
3. Trend-to-Content Translator
Here are 5 industry trends I noticed this week: [LIST]. For each trend, draft a LinkedIn post (under 200 words) that ties the trend to a practical takeaway my audience of [AUDIENCE] can apply this week. End each post with a question that invites comments.
4. Repurposing Engine
Here is a long-form blog post: [PASTE]. Turn it into: (a) a Twitter/X thread of 8-10 posts, (b) a LinkedIn carousel script of 8 slides, (c) a 60-second short-form video script, (d) a 3-email mini-sequence. Keep the hook strong on each version and adapt for the platform's culture, not just its format.
5. Editorial Gap Finder
Here are the last 20 pieces of content I published: [LIST]. Identify three gaps - topics my audience clearly cares about based on the existing themes, but I have not covered. For each gap, propose 3 specific article titles with their primary keyword and the angle that would differentiate them from existing competitor content.
5 Prompts for Ad Copy
6. The Awareness-Stage Facebook Ad
You are a senior direct-response copywriter. Write 5 Facebook ad variations for [PRODUCT] targeting cold traffic that does not yet know they have the problem. Use the Problem-Agitate-Solution structure. Each ad must be under 80 words, lead with a specific moment of pain, and end with a low-friction CTA (read, watch, learn - not buy).
7. The Google Search Ad Set
Generate 10 Google responsive search ad headlines (max 30 characters each) and 4 descriptions (max 90 characters each) for [PRODUCT]. The keyword cluster is [KEYWORDS]. The competitive angle is [DIFFERENTIATOR]. Mix benefit headlines, social proof headlines, urgency headlines, and pure keyword-match headlines.
8. LinkedIn Sponsored Content
Write 3 LinkedIn sponsored content variations for [PRODUCT], audience: [JOB TITLE] at [COMPANY SIZE]. Each post should: open with a counterintuitive insight, reference a specific work scenario the reader recognises, link the scenario to our solution without being pushy, and end with a soft CTA. Word count 120-150.
9. The Retargeting Ad
Write 4 short retargeting ads (under 60 words) for users who visited [PAGE] but did not convert. Address the most likely friction point - [PRICE / TRUST / RELEVANCE / TIMING]. Each ad should reference the specific page they saw and offer a smaller next step.
10. UGC-Style Video Ad Script
Write 3 UGC-style video ad scripts (30-45 seconds) for [PRODUCT]. Format: hook (2 seconds, pattern-interrupt), problem moment (5 seconds, specific scenario), product reveal (5 seconds, natural), proof (10 seconds, specific result), CTA (3 seconds). Voice: real customer, not actor. Avoid marketing language.
5 Prompts for SEO and Keyword Research
11. The Search Intent Mapper
Here is a list of keywords I am considering targeting: [LIST]. For each keyword, identify: (a) search intent (informational / navigational / commercial / transactional), (b) the most likely user question behind it, (c) what content format usually wins on page one, (d) whether I should target this with a blog, landing page, or category page.
12. The Topical Cluster Builder
Act as an SEO strategist. The pillar topic I want to own is [TOPIC]. Build a topical cluster: 1 pillar page, 8-12 supporting articles, and the internal linking structure. For each supporting article, give the target keyword, search intent, and 1-line angle that differentiates it from the top 3 ranking pages.
13. The Featured Snippet Optimiser
Here is my draft article on [TOPIC]: [PASTE]. The primary keyword is [KEYWORD]. Rewrite the introduction and one early section to optimise for featured snippet capture. Use the question, then a 40-50 word direct answer, then expand. Keep the rewrite natural - do not stuff keywords.
14. Competitor Content Gap Analysis
I will paste two competitor articles ranking for [KEYWORD]. Identify: (a) what they both cover, (b) what only one covers, (c) what both miss that the search intent suggests users want. Then propose an article structure that beats both - by being more comprehensive on the missing angle, not by being longer.
15. The E-E-A-T Audit
Here is one of my published articles: [PASTE]. Audit it against Google's E-E-A-T guidelines. Specifically: where does it lack first-hand experience signals, missing expertise indicators, weak authoritativeness, or trust gaps? Give me a 5-item checklist of edits to add genuine experience and expertise without padding.
For an in-depth treatment of writing prompts that actually work, see our guide to AI prompts for beginners.
5 Prompts for Email
16. The Subject Line Generator
Write 20 email subject lines for [EMAIL TOPIC]. The audience is [PERSONA]. Mix four styles evenly: curiosity-gap, benefit-direct, contrarian, and personal-story. All subject lines under 45 characters. After each, give a 1-sentence note on which audience segment it would work best for and why.
17. The Welcome Sequence
Design a 5-email welcome sequence for new subscribers to [LIST]. The lead magnet was [LM]. The eventual offer is [PRODUCT]. For each email give: send timing, subject line, single-idea theme, opening hook, body structure, and CTA. Pace the sequence so emails 1-3 give value, email 4 introduces the offer, email 5 makes the soft pitch with a deadline.
18. The Re-Engagement Campaign
Write a 3-email re-engagement sequence for subscribers who have not opened in 90 days. Email 1: pattern-interrupt subject, acknowledge the silence, ask a real question. Email 2: best-of-our-content roundup. Email 3: clean-list email - explicit ask to stay or unsubscribe, written warmly.
19. The Story-Driven Sales Email
Write a sales email for [PRODUCT] using a story structure. Open with a specific moment from a customer's experience [DETAIL], escalate the tension to the failure point, introduce the product as the turning point, show the resolution, and end with the offer. 250-300 words. UK English. No hype words.
20. The Newsletter Format
I publish a weekly newsletter for [AUDIENCE]. Design a repeatable format with 4 sections that takes me 90 minutes to write each week. For each section: name, purpose, length, content type, and example topic. The format should feel personal, not corporate.
5 Prompts for Audience Research
21. The Persona Sharpener
Here is my current customer persona: [PASTE]. It is too generic. Sharpen it by inferring specifics that change marketing decisions: typical day, what they read, what they fear, what they pretend to know but do not, what they have already tried, what would make them sceptical of [PRODUCT]. Cite the inference logic for each specific.
22. The Voice-of-Customer Mining
Here are 30 customer reviews / support tickets / sales call transcripts: [PASTE]. Extract: (a) the exact phrases customers use to describe the problem, (b) the exact phrases they use to describe the desired outcome, (c) the words they use about competitors, (d) the recurring objections. Give me direct quotes, not paraphrases.
23. The Objection Map
For [PRODUCT], generate the 10 most likely objections a [PERSONA] would raise before buying. For each: the surface objection, the deeper fear underneath, the proof element that addresses it (testimonial, data, guarantee, demo), and where in the funnel to handle it.
24. The Awareness-Stage Mapper
Map my audience by Eugene Schwartz's 5 awareness stages (unaware, problem-aware, solution-aware, product-aware, most-aware). For each stage: percentage of my market in that stage, the language they use, the content that meets them there, and what content moves them to the next stage.
25. The Reddit / Forum Listening Script
I want to research how my target audience [PERSONA] discusses [PROBLEM]. Give me: 8 specific subreddits or forums to check, 12 search queries to run inside them, the patterns to look for, and a structured note-taking template so I can capture insights in a usable form.
5 Prompts for Analytics & Reporting
26. The Monthly Performance Narrative
Here are this month's marketing metrics: [PASTE]. Write a 1-page performance narrative for the leadership team. Structure: top-line summary (3 sentences), what worked and why, what underperformed and why, the single biggest insight, and 3 specific actions for next month. No marketing jargon. No vague phrases like "engagement is up".
27. The Funnel Diagnostic
Here is my full-funnel data: [PASTE]. Identify: (a) the single biggest leak by absolute lost revenue, (b) the leak with the best fix-effort to revenue ratio, (c) the metric that looks healthy but is actually masking a problem. Justify each finding with the specific numbers.
28. The A/B Test Designer
I want to test [HYPOTHESIS] on [PAGE]. Design the experiment: control description, variant description, primary metric, guardrail metric, sample size needed for statistical significance at 95% confidence assuming current conversion rate of [X]%, expected test duration at current traffic of [Y]/week, and what result would change my marketing strategy.
29. The Attribution Storyteller
Here is my multi-touch attribution data: [PASTE]. Tell me the story of how customers actually find and convert with us. Specifically: the dominant first-touch channels, the dominant assist channels, the dominant closing channels, and where I am over- or under-investing relative to the assist value of each channel.
30. The Retention Cohort Reader
Here is my cohort retention data: [PASTE]. Identify: which cohorts retain best and the likely reason (acquisition source, onboarding change, product update), which cohorts churn fastest and the likely reason, and one experiment to improve the worst-performing cohort's retention.
How to Actually Use These Prompts
Three rules.
One: stack them. The biggest gains come from chaining prompts. Use prompt 22 to extract voice-of-customer language, then feed those phrases into prompt 16 to generate subject lines. Use prompt 11 to map intent, then prompt 12 to build a cluster, then prompt 14 to find the gap. The output of one becomes the input of the next.
Two: edit ruthlessly. AI gives you a strong first draft. It does not give you a finished asset. Cut 30% of every output. Add specifics - names, numbers, dates, places - that AI cannot invent. Replace any sentence that sounds like a marketing template.
Three: build a personal brand voice block. Write a 200-word block describing how you write - sentence length, vocabulary register, words you avoid, phrases you use. Save it. Prepend it to every prompt. This is the difference between AI output that sounds like AI and AI output that sounds like you.
Frequently Asked Questions
How do I get AI to write in my brand voice?
Paste 3-5 examples of existing brand-aligned writing into the prompt and ask AI to identify the recurring style traits - sentence length, vocabulary register, rhythm, opinions. Save those traits as a reusable style block and prepend it to every marketing prompt. Refine the block every few weeks as you spot misses.
Which AI tool is best for marketing?
Claude is generally stronger at long-form copy and brand voice. ChatGPT is better for ideation, structured outputs, and tool integrations. For SEO research, Perplexity beats both. Most working marketers use two of the three. See our full comparison.
Can AI replace a marketing team?
No. AI replaces specific tasks - first drafts, research compilation, ad variation generation. It does not replace strategy, judgment, or relationships. A marketing team using AI well will outperform a marketing team of double the size that does not.
Are AI marketing outputs SEO-safe?
Yes, when edited. Google penalises low-quality content, not AI-assisted content. The risk is publishing unedited AI output at scale. The fix is to treat AI output as a first draft, add genuine expertise and original research, and edit aggressively.
How long does it take to get good at marketing prompts?
About two weeks of daily use. The first week feels frustrating. The second week is where it clicks - you start writing prompts that produce usable output on the first try. After a month, your prompt library becomes a real asset.
Want the Full Marketing Prompt Vault?
AI Mastery includes 500+ tested prompts across marketing, sales, content, ops, and personal productivity - all editable, all categorised, all working in 2026.
View Course & Pricing →