One hundred days ago I started a blogging experiment. For the first sixty I wrote almost daily. For the last forty I wrote nothing. This is also ten years since I last wrote anything at all, so the sample is contaminated by every possible variable. Here’s what I think I learned anyway.
The Distribution Reality
I used only LinkedIn to promote the posts — not the full text, just an intro and a link back to the blog. Posts attracted between 50 and 2,000 views on LinkedIn. Roughly 5% clicked through to read the whole thing.
The blog, in other words, is a lonely place. If you’re writing for reach, you’re really writing for LinkedIn and the blog is an archive. I’m fine with that, but it’s worth being clear-eyed about it.
What I cannot be clear-eyed about is which posts perform. There is absolutely no pattern. Posts I was genuinely proud of landed silently. Posts I considered thin and obvious went semi-viral. I have stopped trying to predict this. I suspect nobody can.
The unexpected joy was the comments — strangers I’d never encountered finding the posts and engaging seriously. LinkedIn’s tools for managing this are genuinely terrible, which made it harder than it should have been. But the interactions themselves were the best part of the whole experiment.
What Daily Publishing Actually Does
Daily was too much. Not because of the effort, though it was significant, but because the distribution math doesn’t support it. There’s no compounding effect from posting every day versus three times a week. The algorithm doesn’t reward frequency the way it rewards resonance. I think I could publish weekly and lose almost nothing in terms of reach, while gaining considerably in terms of quality and sanity.
The real cost of daily publishing wasn’t the writing. It was the zone. Getting into the state where the structure of a post reveals itself as you write it — that took time and quiet that was genuinely scarce. Creating the conditions for the zone was more consuming than the writing itself.
Mondays are also, empirically, terrible. Average engagement is roughly half any other day. I don’t know why. I’ve stopped caring why.
The Restart
I’ve been using Claude for a few months now and recently ran a few pilot posts with a different workflow: Claude generates pitches each morning with sources, I study them, push back on the ones I disagree with — which is roughly half — steer the framing, give it my own take, and then ask it to finish the writing from those instructions.
Something unexpected happened. The zone is still here.
I expected the assisted workflow to feel like outsourcing — like the thinking was happening somewhere else and I was just approving it. It doesn’t feel that way. It feels more like talking to someone than writing to myself, and somehow that’s easier to get into flow with, not harder. The focus required to give it good direction, to catch the lazy claims, to say “no, that’s not what I think” — that’s the same cognitive load as writing. Maybe it’s the same act with different mechanics.
The Part That Surprised Me Most
I don’t remember half of what I wrote in those first sixty days.
This bothered me at first. If I don’t remember it, did I mean it? Was it really mine?
I’ve landed on: yes, and yes. Memory and authorship are separate functions. I meant every post when I wrote it. The thinking was real. The positions were mine. The fact that I can’t retrieve them six weeks later doesn’t make them less true — it just means writing is more like speaking than we usually admit. You say the thing, the thing goes into the world, and what you’re left with isn’t the content, it’s the residue of having thought it through.
The AI-assisted posts feel exactly the same. I don’t think I was more “in” the hand-written ones than I am in these. The zone is the zone.
So Why Do We Write
If you don’t remember what you wrote, and you’re not sure which posts will land, and the blog itself gets 5% of the audience of the teaser — why do this?
Two reasons, as far as I can tell.
First, because there are people who want to read short, thought-provoking pieces, and producing those is a way of being useful to strangers you’ll never meet. The comments from people I don’t know are proof of this. It’s a small but real thing.
Second, because you have to self-express. Not to be remembered. Not to build a brand. Just because there are ideas in your head that need to be articulated to be real, and writing is still the sharpest tool for that.
Whether Claude drafts the sentences or you do is, I think, beside the point. The ideas have to be yours. The judgment about what’s true has to be yours. The moment you stop pushing back — stop saying “no, that’s not what I think” — the writing stops being yours, regardless of who typed it.
That’s the only line that matters. I’m still on the right side of it.
Leave a Reply