# ChatGPT Prompt Library for Marketers: 75 Prompts by Role

> 75 copyable ChatGPT prompts for marketers, grouped by role: content, SEO, paid media, product marketing and demand gen.

Source: https://mentionx.ai/learn/chatgpt/chatgpt-prompt-library-marketers  
Author: Dipuck Jones, founder of MentionX  
Updated: 4 November 2026  
Level: Practitioner

**This library holds 75 ChatGPT prompts for marketers, 15 for each of five roles: content, SEO, paid media, product marketing, and demand gen and lifecycle.** Each one names its inputs and the output to return, so you can paste it, fill the brackets and get usable work.

> **Key takeaways**
>
> - Run prompts inside a brand project so voice and facts carry over.
> - Fill every bracket; vague inputs give generic output.
> - Use search or deep research for prompts that need current facts.
> - De-identify any export with personal data before uploading.
> - Turn the prompts you use weekly into skills.

*Checked 4 November 2026 (CET) against OpenAI help center articles and chatgpt.com pages. ChatGPT changes often, so confirm plan details in your own account. "Our view" marks MentionX opinion. Ledgerline is a fictional company used for illustration.*

## How should you use this library?

Each prompt is a starting point. Replace the [brackets], attach the file or paste the text it names, and run it inside your brand project (lesson 3) so it inherits your voice and facts. Prompts that need current facts work best with search or deep research switched on. Anything that touches personal data should use de-identified exports.

## Prompts for content marketers

**Brief from notes**

```
Turn these notes into a content brief: audience, the question answered, our angle, proof, reader next action, out of scope. Ask about gaps. Notes: [paste]
```

**Outline critique**

```
Critique this outline as a busy [role] reader. Which sections would they skip? What question is missing? [paste]
```

**Section from notes**

```
Write the section "[H2]" using only these notes. Answer first, one concrete example, max [n] words, mark unsourced claims [CHECK]. [notes]
```

**Intro cut**

```
Rewrite this introduction so the first sentence answers the headline. Cut anything the reader already knows. [paste]
```

**Claims pass**

```
List every factual claim in this draft and say: supported by the project fact sheet, needs a source, or cut. [paste]
```

**Voice match**

```
Here are 3 published pieces in our voice: [paste]. Describe the voice in 5 rules, then edit this draft to match: [paste]
```

**Cut by a third**

```
Cut this draft by a third without losing any fact or example. Show what you removed. [paste]
```

**Headline options**

```
Give 10 headlines for this piece: 3 question, 3 outcome, 2 number-free how-to, 2 contrarian. Under 65 characters. [summary]
```

**Repurpose**

```
Turn this article into: 1 LinkedIn post, 3 X posts, 1 newsletter section, 5 quote cards. Keep every claim identical. [paste]
```

**Expert questions**

```
I'm interviewing [expert] about [topic] for an article. Write 10 questions that will produce specific stories and numbers, not generic advice.
```

**Case study**

```
Draft a case study from this interview transcript: situation, trigger, what changed, results in their words. Quote only exact phrases. [paste]
```

**FAQ from tickets**

```
From these support tickets (de-identified), list the 10 most common questions and draft a short answer to each from our docs: [paste]
```

**Readability**

```
Mark sentences over 25 words and jargon a [audience] wouldn't use. Suggest plain replacements. [paste]
```

**Update check**

```
This post was written in [year]. List statements that may now be out of date and what I should check. [paste]
```

**Editorial calendar**

```
From these themes and dates [paste], build a 6-week calendar table: date, title, format, channel, owner, goal.
```

## Prompts for seo specialists

**Intent clusters**

```
Group this keyword export by search intent. For each cluster: name, primary keyword, intent, one page or many. Use only the file's volumes.
```

**Cannibalisation**

```
Here are our URLs and their target keywords: [paste]. Which pages compete for the same intent? Recommend merge, redirect or differentiate.
```

**SEO brief**

```
Write a brief for [cluster]: first-paragraph answer, H2s as buyer questions, what we add that ranking pages lack, internal links from [list], meta description under 155 characters.
```

**SERP read**

```
Search for "[query]". Summarise what the top results have in common, what format they use and what none of them cover. Cite each.
```

**Refresh plan**

```
Attached: Search Console queries for [URL]. Which queries does the page rank for but not answer? Suggest new sections and title options.
```

**Title tags**

```
Write 5 title tags under 60 characters for this page that match its top queries: [queries]. Keep the brand at the end.
```

**Meta descriptions**

```
Write meta descriptions under 155 characters for these 10 URLs and summaries: [paste]. Each must state the answer, not tease it.
```

**Internal links**

```
From this sitemap list [paste], suggest 5 internal links for this draft with natural anchor text and the sentence to place them in. [draft]
```

**Schema draft**

```
Draft FAQPage JSON-LD for these questions and answers. Validate that answers match the page text exactly. [paste]
```

**Content gap**

```
Here are Competitor A's blog URLs and titles [paste] and ours [paste]. Which topics do they cover that we don't, grouped by intent?
```

**Redirect map**

```
Map these old URLs to the best new URLs by topic: [old list] [new list]. Flag any with no good match.
```

**Log questions**

```
From these site-search queries [paste], which questions have no matching page? Group them and suggest page titles.
```

**Answer-first rewrite**

```
Rewrite each H2 section so its first sentence fully answers the H2 question. Keep the rest. [paste]
```

**Robots review**

```
Review this robots.txt [paste]. Explain what it allows for OAI-SearchBot, GPTBot and Googlebot. Don't suggest changes yet.
```

**Monthly SEO summary**

```
Attached: this month's and last month's Search Console exports. Summarise clicks, top gainers and losers, and three actions, for a non-SEO reader.
```

## Prompts for paid media managers

**Hook matrix**

```
Write 4 hooks each for pain, outcome, proof, objection and curiosity for [offer] aimed at [audience].
```

**Platform variants**

```
Turn these 3 hooks into full ads for [platform] within [limits]. Show character counts.
```

**Winner patterns**

```
Attached: ad export. Excluding ads under [x] spend, which words, hooks and lengths appear more in the top quartile by cost per conversion? Show counts.
```

**Fatigue check**

```
Attached: daily ad data. Which ads show falling CTR with steady or rising frequency over 14 days? List them with the trend.
```

**Negative keywords**

```
From this search terms report [paste], list irrelevant terms to add as negatives, grouped by theme, with the spend each wasted.
```

**Audience angles**

```
For [product], describe 5 distinct audience segments and the single message most likely to matter to each. Mark guesses [INFERRED].
```

**Landing page match**

```
Here is the ad [paste] and the landing page copy [paste]. Does the page deliver the ad's promise above the fold? Suggest fixes.
```

**Test plan**

```
Plan a 4-week creative test for [budget]: what to test, how many variants, the success metric and the minimum data before deciding.
```

**Budget scenarios**

```
Using this channel data [paste], show what happens to blended cost per lead if we shift 20% of budget from [A] to [B]. State assumptions.
```

**Context hints**

```
Write 10 ChatGPT Ads context hints describing buyer conversations where [offer] is genuinely useful, and 5 where it would be irrelevant.
```

**Video script**

```
Write a 15-second video ad script: hook in the first 2 seconds, one benefit, proof, CTA. Include on-screen text.
```

**Image concepts**

```
Using our style prompt [paste], describe 4 visual concepts for [message], each with a different visual metaphor.
```

**Weekly report**

```
Attached: this week's platform exports. Write a 6-bullet summary: spend, results versus target, best and worst ad, one action.
```

**Compliance read**

```
Check these ads against this list of banned claims and our platform's policy notes [paste]. Flag anything risky.
```

**Retargeting sequence**

```
Write a 3-ad retargeting sequence for visitors of [page] who didn't convert: objection, proof, offer.
```

## Prompts for product marketers

**Positioning options**

```
From these interview notes and win-loss reasons [paste], propose 3 positioning statements with the evidence for each and what's missing.
```

**Messaging hierarchy**

```
Build a messaging hierarchy from [positioning]: core message, 3 pillars, 2 proof points each from this list only: [paste].
```

**Battlecard**

```
From our verified competitor file [paste], write a battlecard for Competitor A: where we win, where they win, landmines, and three questions to ask the buyer.
```

**Launch plan**

```
Draft a launch plan for [feature]: audiences, tier of launch, assets, channels, owners, dates, success metrics. Table format.
```

**Release notes**

```
Turn these engineering notes [paste] into customer-facing release notes: what changed, who benefits, how to use it. No jargon.
```

**Pricing page review**

```
Review this pricing page copy as a sceptical buyer. What's unclear, what would make you leave, what question is unanswered? [paste]
```

**Objection handling**

```
List the 10 most likely objections to [offer] from a [role] and draft a short, honest answer to each.
```

**Persona from evidence**

```
Draft a persona using only this research synthesis [paste]. Quote evidence for each line and mark inferences [INFERRED].
```

**Feature to benefit**

```
For each feature in this list [paste], write the benefit, the proof we have, and the customer phrase that describes the problem.
```

**Sales deck outline**

```
Outline a 10-slide first-call deck for [ICP]: one message per slide, which proof to use, and where to ask a question.
```

**Win-loss themes**

```
Attached: de-identified win-loss interview notes. Group reasons into themes with counts and quotes. What should product, sales and marketing each change?
```

**Review mining**

```
Attached: reviews CSV. Themes with counts and verbatim quotes with row numbers; compare 1-2 star and 4-5 star reviews.
```

**Red team**

```
Act as a sceptical CFO, head of sales and a buyer loyal to Competitor B. Each gives three objections to this plan: [paste].
```

**Pre-mortem**

```
It's six months later and the launch failed. Write the post-mortem: three likely reasons, ignored warning signs, what to change now.
```

**Naming**

```
Suggest 12 names for [feature] in 3 styles (descriptive, benefit, invented). Flag any that may conflict with common terms in [category].
```

## Prompts for demand gen and lifecycle marketers

**ICP from wins**

```
Attached: de-identified closed-won and lost deals. Which attributes and combinations predict wins and short cycles? Propose an ICP with evidence.
```

**Account brief**

```
Research [company] from public sources: business, recent news, relevant hiring, tools mentioned. Cite each point and suggest 2 angles.
```

**Nurture sequence**

```
Write 5 nurture emails over 3 weeks, each answering one buyer question in order. Under 120 words, plain text, one link.
```

**Subject lines**

```
Write 10 subject lines for this email [paste]: 4 direct, 3 question, 3 curiosity. Under 45 characters. No clickbait.
```

**Onboarding emails**

```
Draft a 4-email onboarding sequence that gets a new user to [activation event]. One action per email.
```

**Win-back**

```
Write a 3-email win-back sequence for customers who stopped using [feature] 60 days ago. Ask why in the first email.
```

**Webinar topics**

```
From these sales-call questions [paste], propose 5 webinar topics with a title, promise and three takeaways each.
```

**Webinar follow-up**

```
Write follow-up emails for attendees and no-shows of [webinar], each with one next step.
```

**Lead scoring review**

```
Attached: leads with scores and outcomes. Which scoring rules predict conversion and which don't? Suggest changes with evidence.
```

**Funnel leak**

```
Attached: stage-by-stage counts by source. Where is the biggest drop by source? List possible causes to investigate.
```

**Form questions**

```
Suggest a "How did you hear about us?" question with answer options that include AI assistants, for a B2B demo form.
```

**Event plan**

```
Plan pre-event, at-event and post-event activity for [event]: goals, audience list, messages, follow-up timeline.
```

**Segment messages**

```
For these 4 segments [paste], adapt this email [paste] so the first line and proof point fit each segment.
```

**Re-engagement test**

```
Design an A/B test for our re-engagement email: hypothesis, variant, sample needed, success metric, run time.
```

**Monthly pipeline summary**

```
Attached: pipeline export. Summarise new pipeline by source, conversion by stage and three actions, for leadership.
```

## How do you turn prompts into team assets?

Pick the five prompts your team uses weekly and turn each into a skill (lesson 4), with your templates and approved claims as reference files. Skills are applied automatically when a request matches, so the team gets consistent output without copying prompts around.

*Our view:* a short, maintained library beats a long one nobody updates. Review it each quarter and delete prompts nobody used.

> **Pro move:** add a "last edited" date and owner to each prompt in your team's copy. When ChatGPT's features change, you'll know which prompts to retest first.

## Worked example: Ledgerline's starter set

*Ledgerline is a fictional accounts-payable tool. Details are illustrative.*

Ledgerline's five-person team each picked their three most-used prompts from this library. Seven overlapped, so the team kept eight in total, rewrote them with Ledgerline's templates and built four into skills: claims pass, SEO brief, winner patterns and nurture sequence. The rest stayed in a shared doc.

## What should you check before the next lesson?

- Copy the prompts for your role into your team doc.
- Turn your most-used prompt into a skill.
- The Claude version is the [Claude prompt library for marketers](https://mentionx.ai/learn/claude/claude-prompt-library-marketers).

Next: [Quality, accuracy and brand risk](https://mentionx.ai/learn/chatgpt/chatgpt-quality-accuracy-brand-risk). Or return to the [ChatGPT for Marketers track](https://mentionx.ai/learn/chatgpt/chatgpt-for-marketers).

## Common questions

**Do these prompts work on the free plan?** Most do. Prompts that analyse files, use deep research or use apps depend on your plan's tools and limits.

**Should I paste these into custom instructions?** No. Custom instructions are for rules that apply everywhere. Keep prompts in a doc or turn frequent ones into skills.

**How do I adapt a prompt for my brand?** Run it in a project that holds your voice rules and fact sheet, and replace every bracket with specifics.

## Sources

- OpenAI Help Center, [Projects in ChatGPT](https://help.openai.com/en/articles/10169521)
- OpenAI Help Center, [Skills in ChatGPT](https://help.openai.com/en/articles/20001066)
- OpenAI Help Center, [Data analysis with ChatGPT](https://help.openai.com/en/articles/8437071)
- OpenAI Help Center, [Deep research in ChatGPT](https://help.openai.com/en/articles/10500283)
All accessed 4 Nov 2026.

## Frequently asked questions

### Do these prompts work on the free plan?

Most do. Prompts that analyse files, use deep research or use apps depend on your plan's tools and limits.

### Should I paste these into custom instructions?

No. Custom instructions are for rules that apply everywhere. Keep prompts in a doc or turn frequent ones into skills.

### How do I adapt a prompt for my brand?

Run it in a project that holds your voice rules and fact sheet, and replace every bracket with specifics.

---

Written and reviewed by Dipuck Jones for MentionX Learn. The HTML version of this page: https://mentionx.ai/learn/chatgpt/chatgpt-prompt-library-marketers