I have a name for the way everyone writes on LinkedIn now: engagement speak.
Engagement speak is writing shaped by what the algorithm rewards rather than by what the writer means. The writer starts from a format that gets reach and squeezes whatever they wanted to say into it.
How to Spot It
You know it when you see it. The post opens with a hook, a line built to stop the scroll before it has said anything. Every paragraph is a single sentence, so the post runs tall on a phone. Somewhere in the middle a personal story, a delayed flight or a child’s lemonade stand, turns out to be a leadership lesson. It ends on “Thoughts?” because comments count.
Engagement speak is when it all arrives together and the form matters more than the message.
You see it on LinkedIn, on YouTube, on TikTok, everywhere a feed decides what gets seen. On video it opens with “wait for it” or a shocked face on the thumbnail and closes with a request to like and subscribe. The accent changes from platform to platform. The grammar is the same.
Orwell’s Hen-House
George Orwell described the mechanism in 1946. In Politics and the English Language he complained about prose that “consists less and less of words chosen for the sake of their meaning, and more and more of phrases tacked together like the sections of a prefabricated hen-house.”
Engagement speak is that hen-house with a growth chart attached. The sections are new (“here’s what I learned”, “let that sink in”, “read that again”) but the writer still assembles them from parts someone else cut to size. Orwell’s objection was that ready-made phrases spare the writer the work of thinking. The feed adds a reward for skipping it.
He went further in 1984. The Party designed Newspeak to shrink the vocabulary until certain thoughts could no longer be put into words. Nobody had to design engagement speak. We comply with what the algorithm dictates, mindlessly, one post at a time.
AI as Catalyst
AI is proving a fantastic catalyst for all of this. The language models learned to write from the same feeds, and they reproduce the shapes on request, hook and one-line paragraphs included. Anyone can now turn out a post in engagement speak in seconds, and plenty of people do. Those posts go back into the feed and, eventually, into the training data for the next models.
When It Leaves the Feed
The truly dystopian future is one where all our language coalesces into this. I’m worried it has already started. After seeing it all day, everywhere, I’ve caught myself using these phrases in actual conversation.
At that point the habit has moved out of the writing and into how we talk. Orwell put the risk in one line: “if thought corrupts language, language can also corrupt thought.”
