Hacker News Brand Monitor: Every Mention Of You And Your Competitors

Know the moment your product comes up on Hacker News, in the stories and in the comments where the real opinions live.

A single HN thread can send more qualified traffic than a month of anything else, and the conversation happens whether you are watching or not. Most teams find out days late, when the thread is cold and the correction they wanted to post no longer matters.

Give this a list of terms and get every story and comment that mentions them, newest first, with points, comment count, author, and a link straight to the exact post.

Built for founders watching their own launch, marketers tracking share of voice against competitors, and support teams who would rather answer a complaint on the thread than read about it later.

How it works

flowchart LR
    A[Terms to watch] --> B[Search stories<br/>and comments]
    B --> C[Filter to the<br/>time window]
    C --> D{Does the text<br/>really contain<br/>the term?}
    D -->|no| E[Dropped as<br/>a loose match]
    D -->|yes| F[Merge duplicates<br/>across terms]
    F --> G{New since<br/>the last run?}
    G --> H[(Your mention feed)]

Hacker News search ranks on relevance and splits a query into separate words, so it happily returns threads that never say the thing you asked about. Every result is checked against the actual text before it reaches your dataset. In testing, a search for planet scale returned six results and only one genuinely mentioned it.

What you get

One row per mention. A real row from a live run:

{
  "id": "49290923",
  "type": "story",
  "matchedKeyword": "apify",
  "title": "I built a lint for AI-generated evidence",
  "author": "fmajchrzak",
  "points": 2,
  "commentCount": 0,
  "createdAt": "2026-08-13T19:40:38Z",
  "url": "https://apify.com/...",
  "linkedDomain": "apify.com",
  "hnUrl": "https://news.ycombinator.com/item?id=49290923"
}

Comments come through with the body decoded to plain text, so no ' or <p> tags land in your spreadsheet:

{
  "type": "comment",
  "author": "someuser",
  "title": "Ask HN: What is your testing setup?",
  "excerpt": "if you give them access to Playwright and have them write playwright tests then it works well",
  "hnUrl": "https://news.ycombinator.com/item?id=49610112",
  "storyUrl": "https://news.ycombinator.com/item?id=49609887"
}

Stories carry points and commentCount. Comments carry neither, because Hacker News does not publish a public score for them, so those fields are null rather than a misleading zero.

Running it as an alert feed

Turn on onlyNewMentions and put it on a schedule. The Actor remembers what it has already reported, so each run returns only what appeared since the last one and the dataset becomes an inbox rather than a growing pile.

{
  "keywords": ["yourbrand", "your competitor", "your founder name"],
  "days": 2,
  "onlyNewMentions": true,
  "minPoints": 0
}

Pair that with a daily schedule and a webhook, and a mention reaches Slack while the thread is still on the front page.

Example input

{
  "keywords": ["supabase", "planet scale"],
  "days": 7,
  "includeStories": true,
  "includeComments": true,
  "minPoints": 10,
  "requireExactMatch": true,
  "maxResultsPerKeyword": 100
}

minPoints applies to stories only. A comment has no score of its own, so a points threshold never quietly deletes the comments you came for.

Run it from the command line

curl -X POST "https://api.apify.com/v2/acts/arched_friend~hn-brand-monitor/run-sync-get-dataset-items?token=YOUR_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "keywords": ["yourbrand"],
    "days": 7,
    "onlyNewMentions": false
  }'

Pricing

$0.003 per mention returned. A term with no mentions in the window costs nothing.

	This Actor	Social listening SaaS	Checking by hand
200 mentions a month	$0.60	$99 to $500 a month	Forgotten by week two
Comments included	Yes	Often stories only
Loose match filtering	Yes	Varies
New only mode	Yes	Yes
Data you own	Dataset, API, webhook	Their dashboard

Notes and limits

• Search covers the full Hacker News archive back to 2007, so a long lookback works for research as well as monitoring.
• A single item that matches several of your terms is reported once, with every matched term listed, so it is never billed twice.
• Points and comment counts are read at the moment of the run. A story still climbing will show a lower score than it ends on, so re-running near the end of the day gives the settled number.

Related products

• Brand Mention Monitor for the same watch across news and the wider web.
• Google Reviews Monitor for what customers say where they buy.
• App Store Review Tracker for the same signal inside your app reviews.