Marketing prompt

LinkedIn Post Prompt — Without the LinkedIn Smell

Most AI-generated LinkedIn posts get caught immediately. The tells are predictable: opening question, emoji-bullet list, the phrase ‘Most people don’t realize’, a fake-humble brag, a one-line CTA. This prompt bans every one of those patterns and produces posts that pass for actually-written.

✓ Works in ChatGPT, Claude, Gemini ✓ Tested template ✓ Copy + paste ready

The template

Replace the [bracketed] fields with your specifics. Send to any AI chat.

Turn this idea into a LinkedIn post for [your role].

Rules:
- Hook line on its own — concrete, no questions
- Total 120-180 words
- No emojis
- No "Here's the thing:" / "What if I told you" / "Most people don't realize"
- No emoji-bullet lists (✅ ✨ 🚀 etc)
- One specific example, ideally with a number
- End with a normal closing sentence, not a CTA-y question

IDEA:
[your raw idea — a story, observation, lesson]

How to use it

Replace `[your role]` with how you describe yourself professionally (e.g., ‘product manager at a Series B SaaS’).
Brain-dump your raw idea — it can be one line or a paragraph. The prompt structures it.
Generate, then read it out loud. If you’d be embarrassed saying it to a colleague, edit.

Example output

Here’s an example of what a quality AI returns when you run this template:

A founder I respect told me last month: “Hire slow when you’re growing fast.” I disagreed. Hiring slow when you’re growing fast is how you burn out the team that got you here. The deals that closed this quarter were closed by people working overtime to cover the gaps you should have filled three months ago. The right framing isn’t speed — it’s signal-to-noise. A bad hire in a fast-growing company costs 2-3x what a bad hire in a stable company costs, because the bad hire’s mistakes compound through every team they touch. So: hire as fast as you can find the right people. Slow hiring isn’t a virtue. Patient sourcing is. I’m rethinking how we run our pipeline next quarter — adding two people to do nothing but reach out to candidates we’d actually hire today.

Why this prompt works

Hook line on its own, no question.

Question hooks (‘Want to know my secret?’) are AI-generated LinkedIn’s most obvious tell. A concrete claim with someone’s name in it reads like a real person had a real conversation.

Word count cap fights padding.

Without 120-180, models default to 400-word epics with three section headers. The cap forces editorial choices.

Banned-phrase list targets the specific cliches.

Three years of LinkedIn data shows these exact phrases trigger ‘this is AI’ detection. Banning them by name is more effective than ‘avoid AI tells’.

‘Normal closing sentence, not a CTA-y question’.

AI defaults to ending with ‘What do you think?’ or ‘Drop a comment below’. A simple declarative closing is what humans actually write.

Which AI to use

**Claude** for the prose itself. **ChatGPT** if you want it integrated with web research (looking up a stat to back the post). **Gemini** for posts tied to a Google Doc you’re referencing.

Read the full comparison in ChatGPT vs Claude vs Gemini in 2026 →

Frequently asked questions

Why don’t AI-written LinkedIn posts get engagement?

LinkedIn’s algorithm doesn’t directly penalize AI content, but your audience does. Posts that read as AI-generated get fewer comments and reshares, which reduces algorithmic distribution. This prompt is built to pass the human-detection bar.

Can I have the AI write a whole month of posts?

Don’t. Batch-generated content has structural similarities that readers and the algorithm both detect. One post at a time, edited individually, performs better than 30 posts published from one batch session.

How long should a LinkedIn post be?

120-180 words is the sweet spot for most posts; longer for analytical posts (up to 1,300 words for ‘thought leadership’). LinkedIn truncates at ~210 words on feed view, so your first ~200 words must work on their own.

Should I include emojis?

Industry-dependent. In B2B SaaS and finance, no emojis reads as more authoritative. In creator/marketing/media, light emoji use is fine. The prompt defaults to none — add them back yourself if your niche expects them.

Will my audience know it’s AI-written?

If you edit it carefully and the substance is yours, no. If you publish the raw output unchanged, often yes within the first 2-3 sentences.

Build any prompt in 30 seconds

Free tool, no signup, runs in your browser. Pick a category, fill in the blanks, copy your prompt.

Try the prompt generator → Read the full guide