The HackerNews intent playbook: Show HN, Ask HN, Who's Hiring, and the gold most people miss

HackerNews is a public buying-intent stream that most outbound teams ignore because they think the audience is hostile. The audience is hostile — to spam. Not to context. There's a difference, and the teams that learn it pull more qualified

HackerNews is a public buying-intent stream that most outbound teams ignore because they think the audience is hostile. The audience is hostile — to spam. Not to context. There's a difference, and the teams that learn it pull more qualified pipeline out of HN in a quarter than most SDR pods produce in a year.

We've been reading HN as a signal surface for two years. The pattern is consistent. Four veins of intent, four different reading habits, one rule for engaging. This is the playbook.

HackerNews is hostile to spam. It is extraordinarily friendly to context. The hostility is the filter, not the wall.

HN is an intent surface, not a marketing channel

The mistake teams make on day one is treating HN like Product Hunt. They draft a launch. They time it for Tuesday morning Pacific. They line up upvotes. They get flagged in seven minutes and never come back.

HN doesn't reward broadcasting. It rewards specificity, technical honesty, and useful comments on other people's posts. The marketing-channel framing is the wrong frame entirely. The right frame is: this is a 1.2M-monthly-uniques community where every thread is a transcript of what technical buyers actually care about. You're not there to be heard. You're there to read.

When you read it that way, the math changes. You stop counting upvotes on your stuff and start counting "we're evaluating X right now" mentions in other people's threads. There are dozens per week in any given category. Most go unanswered by the vendors who would obviously be the answer.

Vein one: Show HN — the launchers are the magnet, the commenters are the iron

Show HN is where founders post what they built. The instinct is to look at the launcher and think "is this person a buyer for me?" Usually no. They just shipped a thing. They're in build mode, not buy mode.

The buyers are in the comments. Every Show HN thread that gets traction collects a specific kind of question: "we tried X, does your thing handle Y?" or "I've been looking for something that does Z, are you planning that?" Those comments are buyer-revealed intent. The commenter has just told you, in public, that they have a problem your product might solve. They've also told you what alternative they're using or considering.

We track Show HN comments more aggressively than Show HN launches. The signal-to-noise is better by an order of magnitude. The launcher gets the upvotes; the commenter gets the cold message.

The pattern for engaging: reply in-thread to the commenter (not the launcher), reference their specific question, and give them a real answer — even if the answer is "we don't do that, but X does and here's how." If your product is the answer, a one-line "we built exactly this if you want to see it, no pressure" is fine. If your product is half the answer, say which half.

The thing you don't do: hijack the Show HN itself by replying to the launcher with "great launch — also, we built something similar." That's the move that earns the flags.

Vein two: Ask HN — problems surfaced in public

Ask HN is the most underrated intent vein on the site. People post questions because they're stuck. The question is the problem statement. The thread is them sourcing a solution from the smartest 200 strangers they know.

There are roughly 30–50 Ask HN posts per day that touch a buying-intent topic — "what's the best way to do X," "has anyone solved Y," "we're at $2M ARR and our Z is breaking, what do we use." Filter by "Ask HN:" prefix in Algolia, sort by date, scan for your category language. You'll have a list in 20 minutes.

The reply pattern here is even simpler than Show HN, because the OP literally asked. They want answers. The constraint is that they want answers, not pitches. The shape of a useful Ask HN reply: lead with what you'd do if your product didn't exist, mention your product as one option among others, link to a thing — a doc, a teardown, a comparison — that helps them even if they don't pick you. We walked through the reply mechanics in the Reddit reply guide and most of it transfers cleanly.

Ask HN is a thread of people raising their hand for a solution. If you're the solution and you don't show up, that's a quiet pipeline failure.

Vein three: Who's Hiring — the monthly broadcast of stack and pain

The first weekday of every month, HN runs a Who's Hiring thread. Companies post hiring blurbs. Each blurb contains, on average, three things you can't get cleanly anywhere else: the exact stack they're using, the team's location and size hints, and a description of the problem the new hire will be solving. That problem is, by definition, an unsolved internal pain point.

A "Senior data engineer to help us scale our Snowflake + dbt setup beyond 50TB" tells you: they're on Snowflake and dbt, they're bumping into scale, and they don't have anyone in-house to handle it. If you sell anything in the data observability, cost optimization, or pipeline tooling space, that one blurb is qualification, ICP fit, and pain point on a single line.

The thread runs 1,200–2,000 comments. Manually reading is brutal. The play is to grab the full thread (Algolia gives you JSON), filter by your stack keywords, and surface the 30–80 hits that actually match. Then read those hits. There's more in the Who's Hiring outreach piece, but the short version is: outreach off Who's Hiring works because you're addressing a specific stated need to a specific named team, and they posted the need themselves.

The tradeoff: timing. The thread peaks the first three days, then traffic to it dies. Your reply window is roughly two weeks before the next month's thread eats the attention. Move fast or wait a month.

Vein four: the gold most people miss — comments on unrelated front-page posts

This is the vein nobody talks about. Some of the highest-intent buying signals on HN don't appear in Show HN, Ask HN, or Who's Hiring. They appear as comments on completely unrelated front-page stories.

Someone reads a post about Postgres performance and drops, three layers deep in the comment tree, "we're evaluating Datadog vs Honeycomb vs Grafana right now and the pricing math is brutal." That's a buyer in the wild. They didn't post about being a buyer. They mentioned it sideways while talking about something else. There is no algorithm on HN that surfaces that comment to vendors. Most vendors will never see it.

The cardinality is huge. On any given day, across the front page and the new page, there are dozens of these sideways mentions. We've measured: roughly 60–80% of buying-intent signals on HN appear in this off-topic comment vein, not in the structured Show/Ask/Hiring veins. The structured veins are easier to monitor; the comment vein is where the volume actually lives.

The only way to harvest this systematically is to monitor for category keywords across all comments — not just titles, not just specific threads. Algolia's search-by-date with comment-only filtering plus a tight keyword set gets you most of the way. A buying-signals monitor like Shadow Inbox runs the same query continuously, then puts an LLM intent filter on top so you don't drown in mentions that aren't actually shopping.

How to read HN like an operator, not a reader

There's a difference between reading HN for fun and reading HN for pipeline. The fun version is the front page in the morning with coffee. The operator version is a 25-minute daily ritual structured around the four veins.

The ritual we recommend, in order. First, scan the day's Show HN for category-relevant launches; bookmark any thread with more than 30 comments and come back to read the comments, not the post. Second, run your Ask HN saved search and triage anything new. Third, on the first three days of the month, mine the Who's Hiring thread. Fourth — and this is the part that takes the longest — run your category keyword search across all comments from the last 24 hours and read the top 20 hits.

Twenty-five minutes a day, five days a week. That's two hours and five minutes a week. The output, in a category with reasonable HN density, is 15–40 qualified intent signals per week. We did the reply-rate math in a separate piece; the short version is that even at the low end, that volume of contextual outreach beats sending a thousand templated cold emails by every metric that matters.

Twenty-five minutes a day on HN, done correctly, produces more qualified pipeline than a thousand templated cold emails. The bar is not effort. The bar is reading.

The one rule for engaging

There's one rule, and it has no exceptions. Reply in-thread first. Always.

The reply doesn't have to be a pitch. Most of the time it shouldn't be. It has to be useful — to the OP, to the lurkers, to the search engine. If your reply would be useful to someone who never buys from you, it's a good reply. If it would only be useful if they bought from you, it's spam.

After the in-thread reply, if the OP engages — replies back, asks a follow-up, upvotes — then a DM or email referencing the thread is appropriate. Lead the cold message with the specific thing they said in the thread. Not "I saw your HN comment." That's stalker-coded. Lead with the substance: "you mentioned the dbt scale problem at 50TB — we hit that wall at $REDACTED and ended up doing X. Happy to walk you through it if useful." The thread is your warm intro; the substance is your credibility.

This is the same shape we walked through in the contextual cold message piece. The mechanics transfer across surfaces. The thing that's specific to HN is that the public reply is non-negotiable. Skip it and you're the cold-emailer who showed up at someone's house because they tweeted their neighborhood.

When the playbook stops working

It stops working when you scale it past your ability to be specific. The whole HN model assumes you read each signal, write each reply, and stand behind every claim you make in public. The moment you template the in-thread replies, you're back to spam — and HN catches templated comments faster than email gateways do.

It also stops working if your product doesn't actually fit. HN is full of senior engineers who will read your thing in three minutes and tell you, in public, exactly why it doesn't solve the problem the OP described. That's a feature of the surface. If your fit is fuzzy, you're paying for the fuzziness in public, with your name on it.

The honest constraint: this playbook scales to maybe 50 signals a week per operator before the quality drops. Past that, you need either more operators or a sharper filter. We use an intent-classification pipeline to do the second; most teams should start by doing the first.

● FAQ

Isn't HackerNews hostile to outbound?

It's hostile to spam. It is extraordinarily friendly to context. A reply that demonstrates you read the post, understood the constraint, and brought a useful artifact gets thanked. A copy-pasted pitch gets flagged in minutes. The hostility is a filter, not a wall.

How fast does the front page move?

Roughly 12 hours of meaningful visibility, with a sharp decay after the first 3–4. But comment threads stay live and crawlable for 48–72 hours, and Algolia indexes everything forever. The window for intent is wider than the window for traffic.

What's the best way to monitor HN for intent without sitting on it all day?

Algolia search with saved queries plus the firehose of Show HN, Ask HN, and the monthly Who's Hiring is the manual setup. A buying-signals monitor like Shadow Inbox does the relevance and intent classification on top so you only see posts where someone is actually shopping.

Should I reply in-thread or DM the author?

Reply in-thread first. Always. The comment is your warm intro. If they engage, then a DM or email referencing the thread is fair game. Skipping the public reply and going straight to outbound is the move that gets you flagged.

What about Show HN — those founders aren't buyers, they're sellers.

The founders are sellers. The commenters are buyers. Show HN threads are full of 'we tried X, do you handle Y' questions — that's the intent signal. The launcher is the magnet; the comments are the pile of iron filings.