For five years, I wrote my own LinkedIn posts.
Not always well. Not always on time. But always personally — sentence by sentence, draft by draft, often at 9pm on a Sunday night staring at a blank document.
The output was decent. Sometimes a post would land — 200 likes, a few comments, maybe a DM from someone interesting. But mostly it was a slog. I'd avoid LinkedIn for two weeks, then guilt-post something at midnight, then disappear again.
Six months ago I gave up and let AI write them instead. The engagement went up. Not down.
Here's what actually happened.
Why I resisted for so long
The honest reason: I thought AI-written content felt like cheating. If I let a machine write my posts, was it really me posting? Wasn't the whole point that this was my voice, my ideas, my opinions?
Plus, I'd tried it in 2023. The output was terrible. Every post started with "In today's fast-paced world..." and ended with "What are your thoughts?" The middle was a checkered mess of buzzwords. I couldn't post that under my name.
So I went back to writing manually, posting maybe twice a month, hating Sunday nights.
What changed
Two things.
First, the tools got better. Specifically, the way they handle brand voice. Modern AI marketing tools don't just generate content from scratch — they extract a voice profile from your existing content and apply it to every generation. The output sounds dramatically more like the actual person. (We dug into why this matters here.)
Second, I stopped thinking of it as "AI writes my posts." I started thinking of it as "AI drafts, I edit."
The flow now:
- Tell the tool a topic or idea (sometimes one sentence: "the problem with most onboarding flows")
- AI drafts a LinkedIn post in my voice
- I read it. About 70% of the time, I post it as-is. About 25% of the time, I edit one or two sentences. About 5% I reject and try again.
The whole cycle takes 2-3 minutes per post.
What the data showed
In the six months since I switched:
- Posts published: went from ~6/month to ~16/month (more than doubled)
- Average impressions per post: roughly the same
- Total monthly impressions: increased ~3x (just because I was posting more)
- Comments and DMs: up ~4x
- Profile views: up ~2.5x
- Inbound leads: went from ~2/month to ~7/month
The lift wasn't because the AI posts performed better than my manual ones. They performed about the same. The lift was because I actually posted consistently for the first time in my career.
This is the part nobody talks about with AI content: the consistency benefit dwarfs any quality difference.
What didn't work
The first month was rough.
The drafts felt close to my voice but not quite right. I was editing 50%+ of every post, which barely saved time over writing from scratch.
What helped:
Adding more context to my voice profile. I added specific phrases I use, words I never use, opinions I hold, things my audience cares about. The drafts got 80% closer to my voice.
Letting the tool see my last 20 posts. Most modern tools let you paste in samples or auto-extract from your website. Once it had real samples, the rhythm started matching mine.
Stopping the urge to "polish" every draft. This was the hardest part. I'd want to add a sentence here, tweak a word there, even when the draft was already good. I forced myself to post drafts that were 90% there. They performed fine.
What I learned about voice
The biggest insight: my voice on LinkedIn isn't actually that unique.
I have a few signature moves — short paragraphs, occasional provocation, ending with a thought-provoking question instead of a generic CTA. But 90% of my "voice" is just applied common sense: be specific, don't use jargon, write how you'd talk.
Modern AI can do all of that, given the right setup. It can't replicate my unique observations or specific experiences — but it doesn't need to. I bring those. The AI handles the structure and rhythm.
It turned out the part of writing I dreaded wasn't the thinking — it was the typing. The transition from idea to polished sentence. The AI handles that part beautifully.
The 5-minute weekly system
This is how I do LinkedIn now:
Sunday morning, 5 minutes: I open a notes app and dump 4-6 ideas. Topics, observations, questions. Sometimes from things I read that week. Sometimes from conversations I had. Sometimes just opinions I want to share.
During the week, 2-3 minutes per post: I open the tool, type the topic, get a draft, light edit, post. That's it. The content calendar handles the timing so I don't have to remember.
Total: ~25 minutes per week of LinkedIn. Down from ~2-3 hours when I was writing manually.
The output is more consistent, the engagement is higher, and I no longer hate Sunday nights.
Should you do this?
Probably yes — but only if you handle the voice setup properly.
If you just throw "write me a LinkedIn post about marketing" into ChatGPT, you'll get garbage. The output will sound like every other AI-generated LinkedIn post, and you won't want to associate your name with it. (For a deeper look at why, see our piece on the solopreneur's content stack.)
The trick is investing 30 minutes upfront to set up a real voice profile — your tone, your vocabulary, your forbidden phrases, real samples of how you write. After that, every post takes 2-3 minutes.
Hey Molly handles this voice setup automatically — you connect your domain or paste in 3-5 of your past posts, and the Social Media feature uses that profile for every LinkedIn draft. (Disclosure: I'm one of the people working on Hey Molly. That's why this article exists. But the underlying point — that voice setup matters more than the AI itself — would be true either way.)
The bottom line
The version of me that resisted AI content was protecting something that didn't really need protecting. My posts weren't artisanal. They were mostly observations and opinions, dashed out at 9pm on a Sunday because I felt obligated to post.
Letting AI handle the typing freed me to focus on the part that actually matters: having something worth saying. The result has been more posts, better engagement, more inbound business, and zero Sunday-night dread.
If you're stuck in the same loop — wanting to post more but not having time, or posting inconsistently and feeling guilty about it — try the inverted approach. Write your ideas, let AI draft the prose, and edit only when needed. The free trial covers everything you need.
You might find, like I did, that the version of you who shows up on LinkedIn was never about the typing in the first place.