How to find early users on Hacker News

Most advice about Hacker News is about launching: how to write a Show HN post, when to submit, how to survive the front page. That is a one-time event. This guide is about something you can do every week: finding people on Hacker News who already have the problem your product solves, and being useful to them in the comments.

Hacker News readers are mostly developers, founders and technical people. They are quick to spot marketing and quick to say so. They are also unusually generous with detail about their tools, their workflows and what is not working. If your product is for technical people, or for founders, those details are where your early users are.

Ask HN and Show HN are different places

Two kinds of posts matter here, and they serve opposite purposes.

• Ask HN posts are questions to the community: "Ask HN: How do you manage secrets across environments?" or "Ask HN: What do you use for customer support as a solo founder?". The comments become a long list of real tools and real complaints. These threads are the richest source of people describing a problem.
• Show HN posts are people presenting something they made. The comments are feedback on that product. They are useful for learning what people expect from tools in your space, but replying there to promote your own product is poor form; it is someone else's thread.

Beyond those, regular story threads matter too. An article about a database outage will have comments from people describing their own incidents, and an article about pricing changes at a popular service will have comments from people looking for alternatives. Some of the best opportunities are a few levels deep in a thread about something else.

Search comments for problems, with specific queries

Hacker News has a search box at the bottom of every page. It runs on Algolia, and it lets you search stories and comments separately, sort by date or popularity, and limit results to the past day, week, month or year. For finding people with a problem, search comments, sort by date, and set the range to the past week or month.

Use phrases a frustrated engineer would write. Some patterns that work well, adapted to your product:

• "we ended up building our own" plus your domain, for example "we ended up building our own feature flags". People who built an internal tool are telling you exactly what was missing from the market.
• "is there a tool that" or "is there anything that" plus the job.
• "we switched from" or "we moved off" plus a tool you compete with. The reasons that follow are a list of requirements.
• "the worst part of" plus the task, such as "the worst part of on-call".
• "Ask HN" plus your category, searched as stories, to find whole threads about your problem space.

Put exact phrases in quotes. Try each query with comments only, then stories only, and keep the ones that return people talking about their own situation. Old Ask HN threads are worth reading even when they are too old to reply to: they show you how people describe the problem, which gives you better phrases for next time.

Reply with technical honesty

The community norms are clear: be substantive, be civil, and do not use the site primarily for promotion. A reply that works on Hacker News looks like a comment from a knowledgeable colleague.

1. Address the specific problem with specifics: the approach, the tradeoff, the thing that broke when you tried it. Vague replies get ignored or downvoted.
2. Mention alternatives honestly, including open source and do-it-yourself options. Readers will add them anyway if you leave them out, and it will look like you were hiding them.
3. If your product is relevant, say "I built X, which does this" in one sentence, with the disclosure up front. State the limitations too. "It does not support self-hosting yet" earns more trust than any feature list.
4. Answer follow-up questions directly, including hard ones about pricing, data handling or why someone should not just build it themselves.

What to avoid: marketing adjectives, comparisons that misrepresent competitors, and replying to every thread that loosely relates to your product. A thin reply with a product name in it is likely to be flagged; a detailed reply that happens to mention a product usually will not.

A worked example helps. Say you built a small tool that alerts you when a scheduled job silently stops running. Searching comments for "cron job" "silently" over the past month turns up someone writing that their nightly backup had not run for two weeks and nobody noticed. A useful reply explains the usual causes they should check first, describes the simple heartbeat pattern where the job pings a URL and something alerts when the ping stops, mentions that they could build this with a few lines of script, and only then adds one line: you built a hosted version of exactly this, here is what it does not do yet. That reply is helpful even to readers who never click anything, and it is the kind that is more likely to be received well.

Notice what the reply does not do. It does not claim the product is the only answer, it does not use superlatives, and it does not appear in five other threads that day. On a site where readers can see your entire comment history in one click, consistency is your reputation.

Launching on Show HN is a separate project

A Show HN post can bring a burst of attention, but it is a one-time event with its own rules and its own preparation. This guide deliberately does not cover it. The routine here works whether or not you ever launch, and it often makes a later launch go better: by the time you post, you know how people describe the problem, and a few of them already know who you are.

How often to do this

Hacker News has less volume for any given problem than Reddit or X. Searching every day often turns up nothing new. A pace that works for most products:

• Two or three times a week, run your comment searches for the past few days. Ten minutes each time.
• Once a week, look for new Ask HN threads in your space and read them in full, even if you do not reply.
• Reply to one to three comments a week where you can genuinely add something. On Hacker News, a few excellent replies beat many decent ones.
• Keep participating outside your niche occasionally, so your account history is not only about one product.

Track which queries found the comments you replied to, and which replies led to a conversation. After a month, drop the queries that never produce anything and add variations of the ones that do.

Combining Hacker News with other platforms

Hacker News is rarely the only place your users are. The same person who comments here may ask a longer question on Reddit or complain in a short post on X or Bluesky. The approach is the same everywhere: listen for problems, not product names, and reply as a person. The social listening guide for indie hackers covers how the platforms differ, and the Reddit guide covers the platform with the most volume.

Running the searches by hand across several platforms is where the routine usually breaks down. Ownfeed does that part: you paste your product URL, it searches public posts on Reddit, X, Bluesky and Hacker News every day, and AI keeps only the posts where your product would help. You still read them and write every reply yourself; it never posts for you.