Asset 20 8 2
Does AI recommend your business? Run the free check →

Join 15,000 business owners, marketers and entrepreneurs. The Sunday newsletter you'll be annoyed only arrives once a week.

Article

Grok Prompts for X Posts That Get Engagement

If you are skim reading
The short version: Grok writes punchier X posts than ChatGPT when you give it a strong opinion and a specific audience to argue with, not a generic topic. The prompts that get replies tell Grok who to annoy, not just what to say.

The short version: Grok writes punchier X posts than ChatGPT when you give it a strong opinion and a specific audience to argue with, not a generic topic. The prompts that get replies tell Grok who to annoy, not just what to say. Below are the exact prompts I'd use, a worked example, and the uncomfortable bit nobody mentions: engagement prompts don't fix a boring account.

Why Grok is different from ChatGPT for X posts

Grok sits inside X, pulls in live trends, and has a looser, more argumentative tone baked in by default. That's useful and dangerous at the same time. Useful because it can reference what's trending right now without you feeding it context. Dangerous because its default output leans snarky, and snark without substance reads as try-hard within about three posts.

If you've turned the Grok button off because it kept popping up uninvited, fair enough, I've written about how to turn the Grok button off for people who want it gone from their timeline but still use it deliberately through grok.com or the API when they choose to. That distinction matters. Ambient Grok suggestions are mostly noise. Deliberate, well-prompted Grok sessions are a different tool from ChatGPT for this specific job.

The prompt structure that works

Most people prompt Grok like they'd prompt ChatGPT: "write a tweet about [topic]." That gets you generic output with Grok's voice pasted on top, which is worse than plain generic because it sounds like it's trying to be edgy and failing.

The structure that works has four parts, in this order:

  • A stated opinion, not a topic ("most small businesses waste their ad budget on boosted posts" not "social media ads")
  • Who you're disagreeing with (a type of person, not a named individual)
  • The format constraint (character count, thread vs single post, whether you want a question at the end)
  • A ban list of clichés you don't want (this is the part almost everyone skips)

Here's what that looks like in practice:

"I run a small accountancy practice. Write me 5 single X posts, under 260 characters each, arguing that most freelancers underprice because they're scared of losing clients, not because the market is competitive. Write to annoy freelancers who brag about being 'fully booked.' No emojis, no 'hot take,' no rhetorical question at the end, no hashtags."

That prompt gives Grok an opinion, a target, a format and a cliché ban in one go. Compare the output to a vague prompt and the difference is immediate: you get posts with a spine instead of posts that could have been written about any topic with the nouns swapped out.

A worked example, start to finish

Say you run a small B2B software company selling inventory tools to independent garden centres. You want X posts that get replies from other small retail software founders, not just likes from people who'll never buy.

Step one: feed Grok the opinion, not the product. Prompt: "I sell inventory software to garden centres. Write me a post arguing that most retail software fails not because of bad features but because founders build for retailers they've never worked a till for. Make it punchy, under 280 characters, no hashtags."

Step two: ask for three tonal variants of the same point, so you can pick the one that sounds like you rather than like Grok. Prompt: "Give me three versions of that post: one blunter, one more self-deprecating, one that ends with a direct question to other small software founders."

Step three: ask Grok to stress-test the claim before you post it. Prompt: "What's the strongest counterargument someone in retail software would make against this post? Give me one line I could use as a reply if someone pushes back."

That third step is the one almost nobody does, and it's the one that stops you posting something that gets torn apart in the replies by someone who knows the sector. It also gives you a ready-made reply, which keeps the thread going, which is most of what X's algorithm rewards.

Work with me

Want AI doing the heavy lifting in your marketing?

I build the systems that handle the boring 80 percent, so you get your week back. Done properly, with the human kept in.

Step four: post at a time when your actual audience is on, not when a generic "best time to post" list says. For most B2B accounts that's mid-morning UK time or early afternoon US Eastern, but check your own analytics rather than trusting a blog post's guess, including this one.

Ten prompts worth keeping on hand

  • "Turn this paragraph from my newsletter into a 3-post thread with one clear argument per post."
  • "Write a post that disagrees with the common advice to [X], aimed at people who've been burned by following that advice."
  • "Give me a one-line reply to this post that adds a fact, not just an opinion." (paste a competitor's post)
  • "Rewrite this post to remove anything that sounds like marketing copy."
  • "What's trending in [your industry] on X right now that I could have a genuine opinion about?"
  • "Write this as a question that a beginner in [industry] would ask, not a rhetorical one."
  • "Give me a contrarian take on [trending topic] that I could defend if challenged."
  • "Shorten this to under 200 characters without losing the specific number in it."
  • "Write a reply to someone who disagrees with my post, firm but not defensive."
  • "List the three most overused phrases in my niche's X posts right now so I can avoid them."

Notice how many of those ask Grok to cut or defend rather than generate from scratch. That's deliberate. The generation step is the easy part; the editing and stress-testing steps are where engagement gets built.

The uncomfortable bit

None of these prompts will save an account that has nothing to say. Grok can sharpen a weak opinion into a sharper sentence, but it can't manufacture an opinion you don't have. I've seen accounts (and I mean the pattern generally, not one specific case) pump out a dozen Grok-polished posts a week, all technically well-formatted, all forgettable, because the person behind them hasn't decided what they think about their own industry. The prompt isn't the bottleneck. The opinion is.

If you're using AI tools across your whole content calendar and not just for one-off posts, it's worth looking at the wider toolkit rather than treating Grok as a standalone trick; I've covered AI powered tools that can help you manage social media if you want the broader picture. And if ChatGPT is still doing most of your heavy lifting for captions and content ideas, my list of ChatGPT prompts for social media managers covers the prompts I use week to week, Grok just slots in alongside those rather than replacing them.

Where Grok's trend awareness pays off

The one genuine edge Grok has is live context. Ask it "what's being argued about in [your niche] on X today" and it will often surface a real, current argument you can jump into, rather than you guessing at a trend from three days ago. That's worth using for timing, less so for tone. Use Grok to find the argument, then write the actual post in your own voice, or edit Grok's draft hard until it sounds like you and not like a debate bot.

If you're deciding which part of X to even focus on before you bother prompting anything, it's worth reading my take on the best Twitter niches for business growth and the ones I'd avoid, because the prompt only matters once you've picked a lane worth shouting in. Same logic applies if you're sizing up what competitors in your space are saying, which is where a structured competitor analysis prompt set earns its keep before you start drafting anything.

Frequently asked questions

Does Grok write better X posts than ChatGPT?

For X specifically, often yes, because Grok has live trend context and a looser default tone, but it needs a strong opinion and a named audience in the prompt or it just sounds snarky without substance.

What's the single biggest mistake people make prompting Grok for engagement?

Asking it to write about a topic instead of giving it an opinion to argue and a type of person to argue with; vague prompts produce vague, forgettable posts no matter how good the tool is.

Can Grok prompts fix a low-engagement X account?

No, they can sharpen an existing opinion into a punchier sentence but they can't create an opinion you don't have, and that's usually the actual problem behind low engagement accounts.

How many posts should I generate per Grok session?

Ask for three to five variants of the same core point in different tones, then pick the one that sounds most like you rather than posting the first draft.

Run a business, a brand or a blog? I feature tools, products and expert guest posts for an audience of 15,000. See how it works

Published and maintained by the Lilach Bullock team, covering marketing, AI and business growth.
Your buyers are asking AI who to use. Does it say you?

See for free whether ChatGPT, Claude, Perplexity, Gemini and Google name you, and get the plan to become the answer.

Check my AI visibility →
Sundays only

Get the Sunday newsletter.

One email a week. AI experiments, marketing tactics, and the workflows Lilach is building right now in her own business.

Subscribe free

Let’s get your marketing running on AI.

Book a free 30-minute call

We figure out what you need, where AI fits in, and what working together would look like.

Book the call →

Or take the 30-second calculator

You’ll see the hours and the money quietly leaking out of your week, and the three workflows worth building first.

Take the calculator →

Or grab the free AI resource library

Prompt packs, templates, checklists, and swipe files. The exact tools I build for paying clients. Yours, free.

Get the library →
Keep reading

More from the blog.