The Hacker News launch playbook
Front page means thousands of users by lunch. Page three means two visitors, one of them you. The variance is why it's worth doing right.
Why HN is different from every other launch channel
Hacker News has no algorithm to game, no paid tier, no homepage takeover to buy. Stories rise on early upvotes and decay on flags and age — the whole mechanism fits in one page of code. That means the bar is brutally honest: the post either makes a skeptical, technical, hype-allergic audience want to click, or it dies in /newest in twenty minutes.
The audience is also uniquely valuable for dev tools and AI products: they're exactly the early adopters who install extensions, try CLIs, file real bug reports, and leave detailed comments that are better user research than most startups commission. A front-page Show HN isn't just traffic — it's a hundred qualified product reviews for free.
The catch is the variance. Most Show HNs die quietly — that's not an insult, it's the base rate. The playbook below is about moving the odds, not guaranteeing the outcome.
Key takeaway: HN is the one channel where being technically excellent and honest beats being well-marketed. Play to that.
What survives /newest
The posts that make it share a pattern. The title does the work: 'Show HN: I built X that does Y' where Y is a concrete outcome, not a category. 'AI replies in your voice' beats 'an intelligent messaging platform.' No exclamation points, no 'revolutionary,' no emoji — HN flags marketing language on sight.
The first comment matters as much as the title. Post a comment from your own account the moment it's live: what you built, why, what's interesting technically, and what you're unsure about. HN rewards founder-presence — the posts that thrive are the ones where the builder is in the thread answering questions honestly, including the hard ones.
Timing helps at the margins: US morning weekdays catch the most /newest reviewers. And whatever you do, don't ring upvotes — HN's ring detection is excellent and a flagged launch is worse than a dead one.
Title = outcome, not category
'Show HN: Compress video 10x in the browser' outperforms 'Show HN: Next-gen video platform' every time.
Founder comment immediately
Context, honesty, the interesting technical detail — it sets the thread's tone before the cynics do.
Answer everything, especially critics
The skeptical questions are the free user research. Handled well they become the best part of the thread.
Never buy or beg votes
Ring detection kills the post and the account's future submissions. Two organic upvotes beat twenty summoned ones.
Key takeaway: Write the post for the /newest reviewer deciding in three seconds — concrete outcome, zero hype.
Whether the post hits or not, bank the byproducts
A Show HN that dies at 4 points still did its job if you collect what it produced: the comments (even three are real user feedback), the submission's permanent backlink from one of the web's highest-authority domains, and the data point about whether your headline made people click. Founders who treat it as all-or-nothing miss that HN pays even when it misses.
And one dead Show HN doesn't close the door — reposts are allowed after a while if the product meaningfully changed. Ship the feature the commenters asked for, then try again with the better story. The second attempt with real iteration reads completely differently than a relaunch.
Stack it with the permanent layer: the listings and directory pages keep working regardless of what HN's RNG did that morning.
Bank the permanent part first
Before you gamble on /newest, get the listing that works regardless — free and live in minutes.
Related articles
Building something? Tell the story.
Write about what you're shipping — posts are indexed, linked to your product, and read by people looking for new tools.
Write a post