Two different machines
| Hacker News | X | |
|---|---|---|
| Surface | One ranked front page, the same for everybody | Personalised feeds, no single front page |
| Decided by | Upvotes in roughly the first hour or two, against a time decay | Engagement per impression, tested in expanding rings |
| Moderation | Human, attentive, and willing to penalise a domain permanently | Automated, at the scale of the platform |
| Shape of the day | A sharp spike while you hold a front-page slot, then it ends | A curve over roughly 24 to 30 hours, with a tail |
| Second chances | Rare. Moderators occasionally re-surface good posts that missed. | Real. Second and third waves carry past the first-day decay. |
| What you keep | Traffic, and a thread that ranks in search for years | Followers, reachable again for free |
The row that matters most is the last one, and it is the reason this comparison rarely resolves into picking one.
Traffic versus audience
A front-page slot on Hacker News is among the best free traffic events available to a technical product. It is also almost entirely anonymous and non-repeatable. Thousands of people arrive, some try the thing, and you have no way to reach any of them again except by earning another front page.
An X launch usually produces a smaller same-day number and leaves behind followers on your own account, plus a raised baseline that the next launch starts from. That is the compounding difference between a channel that pays once and one that pays into the next thing you do.
Neither is better. They are different assets, and a company that only ever does the first one is repeatedly re-earning an audience it could have kept.
What Hacker News rewards
The community is engineers and technical founders, heavily concentrated in the US and Europe, with real evaluation influence over developer tools and infrastructure. It is smaller than X and considerably denser.
- Something you can try in thirty seconds. The single strongest predictor. A working demo with no account required outperforms everything else, and the closer the first interaction is to the actual product the better it goes.
- Technical depth, published rather than promised. An architecture write-up, a benchmark with methodology, the hard problem and how you solved it. The audience wants the engineering, and giving it to them is the marketing.
- Honesty about limitations, stated by you before somebody finds them. "This does not work for X yet, here is why" buys more goodwill than any claim, because it signals you know where the edges are.
- The founder in the comments. Answering everything, for hours, including the harsh ones. This is the highest-leverage six hours in the whole launch, and the thread's tone is usually set inside the first three replies.
- A plain title. "Show HN: a profiler for Rust async code" beats anything with an adjective in it.
Timing is worth a small amount of thought and less than people give it. Weekday mornings US time is the conventional slot, on the reasoning that the front page is decided by whoever is awake in the first hour.
It is a real effect at the margin and it cannot rescue a post that gives nobody a reason to click.
What it punishes
- A signup wall. The fastest way to lose the thread. A waitlist or mandatory account turns the top comment into a complaint about the waitlist, and the top comment frames everything below it. If you are launching a waitlist, this is the wrong venue for it.
- Marketing language. The register that works on a landing page reads as evasion here. Every adjective in the title is a small tax.
- A landing page where a product should be. Show HN carries an expectation that there is a thing. A page describing a thing is a different kind of post and gets treated as one.
- Arguing with critics. Defensiveness in the replies is more damaging than the criticism, because the whole audience is reading how you handle it. Concede the fair points quickly and answer the rest with specifics.
- Anything resembling vote manipulation. Covered below, and the consequences are the most severe on this list.
The part nobody can sell you
Soliciting votes is against Hacker News rules. The moderation there is human, unusually attentive, and the penalties run from a post being silently deprioritised to a domain being banned permanently. That last one is not theoretical and it is not reversible on a timeline that helps you.
So: we do not offer Hacker News amplification, and we would advise against anyone who does. There is no version of a creator network that applies to a community explicitly built to detect and punish exactly that.
Asking your team, your investors or your users to go and upvote is the same rule and the same risk with a friendlier framing.
What is left is preparation, and it is genuinely all there is: have something worth trying, write a plain title, publish the technical detail, and be present in the comments for the day. That is the entire strategy, and the reason it reads as thin is that the venue is designed so it should be.
Running both on the same day
They interact well when sequenced and badly when merged.
- Post to Hacker News first, weekday morning US time, and clear your calendar for the comments. It is the one that needs you personally and the one you cannot influence any other way.
- Run the X launch the same day, on its own material. The claim, the demo, the number that surprised your team. Amplified distribution belongs here, where paid amplification is allowed and expected.
- Do not post the Hacker News link on X asking for support. This is the specific move that converts a good day into a domain penalty. Sharing that a discussion is happening is different from asking anyone to go and vote in it, and the difference is worth being careful about rather than clever about.
- Let the outcome feed the next post. A strong HN thread often contains the best objection anyone raised about your product. Answering it publicly is a better second-wave post than anything in the original plan.
Which fits what you are launching
| If you are launching | Where the weight goes |
|---|---|
| A developer tool, CLI, library or open source project | Hacker News first, X the same day. The demo does the work in both places. |
| Infrastructure with a benchmark to show | Both, evenly. The methodology post is the asset and it travels in both venues. |
| An AI product with a visible output | X leads. The output is the post, and there may be nothing to try in thirty seconds. |
| A funding round | X only. Funding is not a Hacker News story and posting it as one goes badly. |
| A waitlist or an early-access programme | X only, and reconsider the waitlist. |
| Anything enterprise, sold through procurement | X, plus the artifacts. The shortlist is written elsewhere. |
Product Hunt is a third option with its own rules and its own honest assessment, covered in the Product Hunt guide. The short version is that it suits a different audience again, and that all three can run in the same week without any of them cannibalising the others.
Questions we get asked
Can you run a Hacker News launch the way you run X?
No, and nobody legitimate can. Soliciting votes is against Hacker News rules, the moderation there is human and attentive, and the penalties reach as far as a permanent domain ban. We do not offer it, we advise clients against anyone who does, and the only preparation that works is having something genuinely worth trying.
Which is better for launching a developer tool?
Run both, because they pay differently. Hacker News delivers the best single-day traffic available for free to anything with a demo somebody can try in thirty seconds. X delivers a smaller same-day number and leaves you followers who are reachable again for nothing, which compounds into the next launch.
What kills a Show HN post fastest?
A signup wall. The community's expectation is that they can try the thing immediately, and a waitlist or a mandatory account turns the top comment into a complaint about the waitlist, which then frames the entire thread. Marketing language in the title and a founder who argues with critics are the next two.