The short answer: AI is useful for LinkedIn in a narrow way — sharpening the hook, tightening a draft you wrote, and generating angles when you have nothing. It is actively counterproductive if you let it write the whole post, because LinkedIn audiences are unusually sensitive to the generic register, and a formulaic post costs you credibility with exactly the people you are trying to reach.
| Part | Length | Notes |
|---|---|---|
| Hook | First 2 lines (~210 characters before "see more") | Everything depends on this |
| Body | Usually 800-1,200 characters | One idea, told concretely |
| Close | 1 line question or takeaway | A question invites replies, which extend reach |
| First comment | Optional | The usual place for links |
Why links go in the first comment
Posts containing outbound links tend to get less distribution, so most people put links in the first comment and say so in the post. This is not a trick that overcomes bad content — it just stops the link from working against a post that would otherwise do fine.
Dwell time beats length arguments
Much of the advice about ideal post length misses the mechanism: the platform responds to how long people actually spend with your post. That means line breaks, a genuine hook, and a body that delivers on it. Ask AI for structural help — where to break lines, whether the first two lines work as a hook — rather than asking it to hit a word count.
What AI is actually good for here
- Hook variants. Ten versions of your opening two lines is the highest-value request.
- Cutting your own draft. LinkedIn posts improve with almost every deletion.
- Finding the idea. Paste a messy incident from your week and ask what the generalisable point is.
- Repurposing. One detailed post becomes a comment, a short note and a talk track.
What gets you ignored
- Invented stories. Fabricated anecdotes about employees, clients or strangers are common, detectable, and damaging when noticed.
- The same cadence as everyone else. Short lines, a dramatic turn, a moral at the end — readers have seen hundreds.
- Broken-English inspirational closings. The format is so recognisable it reads as automated.
- Precise fake numbers. "I increased revenue 312% in 30 days" invites scrutiny you cannot survive.
A working process
- Start with something that actually happened. A decision, a mistake, a number you can defend.
- Write it badly first. Get the substance down without editing.
- Use AI to tighten, not to generate. Ask it to cut weak lines and flag anything unsubstantiated.
- Rewrite the hook last. Once you know what the post says, writing the opening is much easier.
Get the prompts
Prompts for hooks, tightening drafts and repurposing one post into several formats.
Get Free Prompts →What do people ask about AI writing tools for LinkedIn posts?
Can AI write LinkedIn posts?
It can help with hooks, tightening your own draft and generating angles, but writing the whole post usually produces the generic register LinkedIn audiences distrust. Use it as an editor rather than a ghostwriter.
Should I put links in the post or the comments?
The first comment. Outbound links in the post body tend to reduce distribution, though this helps only if the post itself is worth reading.
How long should a LinkedIn post be?
Usually 800-1,200 characters, but length matters less than whether people stop and read it. Focus on the hook and line breaks rather than hitting a word count.
What gets posts ignored on LinkedIn?
Fabricated stories, the standard dramatic-cadence format, motivational closings, and precise numbers you cannot defend if someone asks about them.