Content3 min read

X for developers who are not trying to be influencers

The format rewards one sharp observation and punishes everything else. That suits technical people better than it first appears.

X looks hostile to technical content. Short format, fast feed, and a culture of engagement bait that makes most developers want nothing to do with it.

But the actual mechanic, one idea stated sharply, suits engineers well. The obstacle is mostly the surrounding noise, which you can ignore.

The format is one idea

280 characters holds one thought. Not one thought plus context plus a caveat. One thought.

That constraint is the whole thing. A post that tries to do two jobs does neither, and the reason most technical posts fail on X is not that they are too technical. It is that they are trying to include the setup.

Fixed a race condition in our scheduler. It only triggered when two workers polled within the same millisecond, which happens more often than you would think at 15 minute intervals.

That is one idea with a specific detail. It fits. It does not need the architecture explained.

What works

A single observation. Something you noticed that someone else might not have.

A shipping update with one detail. Not a changelog. One specific thing about what you shipped.

A decision and its reason. "Went with X over Y because Z." Compresses well and is genuinely useful.

A finding. Something that surprised you. Surprise is the most reliable interest signal there is.

What does not

Preamble. "So I've been working on something and wanted to share..." You have spent a quarter of your characters saying nothing.

Thread bait. "A thread 🧵" as the opening. If the thread is good, the first post should stand alone and people will keep reading.

Hashtag blocks. Two at most. Usually zero. Hashtags on X are a relic and stuffing them reads as spam.

Filling the limit. 280 is a ceiling, not a target. A 120 character post with a sharp idea outperforms a 279 character post with a diluted one.

The relationship to LinkedIn

Do not shorten a LinkedIn post to fit X. That produces something wrong for both.

LinkedIn rewards a considered story, so the payoff can come at the end. X rewards a live conversation, so the payoff has to come first. Trimming a LinkedIn post keeps the setup and cuts the payoff, which is exactly backwards.

The X version of the same work is restructured, not shortened. Frequently it is a different angle on the same material entirely.

Frequency

X tolerates more volume than LinkedIn because the bar per post is lower and the feed moves faster. Daily is sustainable if you have the material.

The trap is treating that as an obligation. Posting daily with nothing to say produces a feed of thin observations, and thin observations are worse on X than anywhere else, because there is no length to hide behind.

Replies are underrated

Most of the value on X for a technical account is in replies rather than posts.

A useful reply to someone with a larger audience reaches more people than your own post will, and it is a lower barrier: you are responding to something specific rather than generating from nothing.

Good replies add information, offer a counterexample, or share a relevant experience. Bad ones say "this" or promote your product. The difference is whether the reply would be worth reading by someone who is not you.

xtwitterwritingdevelopers

Keep reading