Why Your Last Quarter of Posts All Said the Same Thing

Open your own LinkedIn profile and read your last twenty posts in one sitting. Not skim — read them, in order, the way a prospect does when they're deciding whether to take you seriously.

Most founders who do this get the same surprise. The posts aren't bad. Individually they're fine. Read together, they're one message wearing twenty outfits.

This is the failure nobody writes about, because it is invisible at the only altitude anyone inspects. You review your posts one at a time; the sameness lives in the body of work. It's also not a failure of effort or imagination — it's the predictable arithmetic of how the posts were produced.

Why it happens, mechanically

You have one value proposition. It is the clearest, most rehearsed thing you own — you've said it on calls, put it on the website, pitched it to investors.

In a marketing team this problem has a familiar shape: a quarter's posts all hung off the same launch or the same anchor asset. The launch is the source, so everything converges on it, and the convergence looks like discipline while it's happening. A founder's version is worse only because the single source is permanent. A launch ends.

So when you sit down to write, that's the object you generate from. Post one is the value proposition with a story attached. Post two is the value proposition framed as a problem. Post three is the value proposition as a lesson learned. Each of them looks like a different post while you're writing it, because each has a different opening and a different anecdote.

They are all paraphrases of the same sentence.

Nothing about that is lazy. It's what happens when a single source is narrowed the same way every time, and nothing in the writing process ever compares this post to the last nineteen, so nothing ever catches it. AI writing tools accelerate the problem rather than fixing it, because the value proposition is precisely the thing you put in the system prompt.

The two narrowings

Every good post is narrow, and the fix has nothing to do with writing vaguer ones. The problem is where the narrowing happens.

The bad narrowing happens once, globally, at the source. You decide the company is about one thing, and every post starts from that one thing. Each post then narrows further to a single point — but it was already narrow before you began, so the twenty points are all within a few degrees of each other.

The good narrowing happens per post, from a wide pool. Each post pulls from a rich set of possible material and narrows that to one sharp point — and because the pool is wide, different posts land on genuinely different points.

Both produce the same surface: one clear message per post. The machine underneath is the opposite. And the difference is only ever visible in the corpus, which is the one thing nobody reviews.

The objection, which comes from people who do this for a living

When I ran marketing teams I wrote posts five at a time, deliberately, for continuity of message. The reasoning is sound and I'd defend it now: most people will see one of the five, if that. You don't want to be one thing on Monday and a different thing on Thursday. Consistency of position is not a failure — it's most of what a brand is.

So which is it?

Both, and the answer is the two narrowings again. Continuity of message means the position holds across the five. Convergence means the five posts are the same post. You can have the first without the second: five angles, five vantages, five pieces of source material, all pointing at one position. That's a coherent quarter, and it's what the discipline is for.

What you can't do is manufacture continuity by writing the same post five times. That's what happens when the position stops being the destination and becomes the source.

The same logic explains why anchors are fine and permanent anchors aren't. The stronger the asset, the tighter everything bends toward it — a Gartner Magic Quadrant reprint is expensive enough that you'd push it hard for three to twelve months, and everything in that window pointing at it is correct behaviour. It ends. Your value proposition doesn't. A permanent anchor made of your own pitch is the same mechanic with nothing behind it and no expiry date.

What a wide pool is actually made of

If the fix is a wider source pool, it's fair to ask what goes in it. Four things, and only one of them is your value proposition.

Angles. A theme should expand into three to five declared angles — distinct claims that all point at the same position without restating it. Not “we help small teams market,” but the several separate things you believe that add up to that. Written down, so you can see which you've used.

Vantages. The same founder can write from more than one standing position: as the person running the company, as an observer who has watched this pattern many times, from the customer's perspective, or from the specialist's. These are not different authors — on your profile the speaker is always you. A customer-vantage post centres the customer's experience, told by you. It never ghost-writes as them, which would be both dishonest and immediately detectable.

Sources. This is the biggest lever and the most neglected. A post generated from your value proposition is a post about you. A post generated from a source — a customer conversation this week, something you read that's wrong, a competitor's claim you disagree with, a piece of your own work, a change in your market — is a post about the world, with you in it. The second is more interesting to read and structurally harder to converge, because the world supplies new material and your value proposition doesn't.

At least one post per window that doesn't sell. Something that lands near your theme without arguing for you. If every post sells, the corpus reads as a brochure regardless of how good each post is — and a reader who has classified your feed as a brochure has stopped reading it as evidence.

The check, and why nobody runs it

Convergence cannot be detected by reviewing the post in front of you. By construction. The post in front of you is fine.

It can only be detected by reading a quarter of your own output in one sitting, deliberately, asking a specific question: how many genuinely distinct claims are in here? Not topics — claims. Twenty posts frequently resolve to three.

Almost nobody does this, for a reason that isn't laziness either. Re-reading your own marketing is unpleasant, produces nothing, and has to be scheduled by you against no deadline. It's the same structural problem as the quarterly framework in the previous post: high value, zero urgency, no forcing function.

I've written several hundred LinkedIn posts and I have never once sat down and read a quarter of them back. I'm not offering that as a confession. I'm offering it as evidence of how little the professional version of this job forces the check either.

And you have to do it against a written record of which angles and vantages you've already used, or you're relying on memory of things you wrote weeks ago while thinking about something else.

A worked example, since this series is itself eight posts on one subject — the exact conditions under which convergence happens. Before any of them were drafted, each was assigned an angle and a vantage, and the assignments were written down. The eight LinkedIn posts announcing them were assigned separately, because publishing eight structurally identical announcements would have demonstrated the failure this post describes. That's the whole discipline. It's clerical, it takes about twenty minutes, and nothing whatsoever would have gone wrong on the day if we'd skipped it.

Why this is the expensive one

The other failures in this series are visible. A post with AI tells in it looks wrong on the page. An inconsistent cadence is a gap in the feed anyone can see.

Corpus convergence is invisible to the person producing it and obvious to the person consuming it. Your prospect reads twenty posts in five minutes and forms one impression: this company has one thing to say and says it constantly. That impression is nearly impossible to reverse, because it was formed by the totality of your work rather than by any single piece you could go back and fix.

It also punishes effort. The founder who posts twice a week for six months builds a larger, more visible demonstration of sameness than the one who posts occasionally. Doing more of it makes it worse.

What this costs to do properly

Stated honestly, as a standing obligation rather than a tip:

1. A theme expanded into three to five named angles, written down at the start of the quarter.
2. A record of which angle and which vantage each post used, maintained as you go.
3. A habit of pulling from sources rather than from your value proposition — which means having somewhere real material accumulates, and reading it before you write.
4. At least one non-selling post per window, deliberately placed.
5. A quarterly read-through of everything you published, against the question of how many distinct claims survived.

None of it is difficult. All of it is ongoing, and every item is the kind of task that loses to a customer email.

There's a second half to why a corpus feels repetitive, and it has nothing to do with what you said. It's how it sounded.