How to use HN's Who's Hiring thread for cold outreach that doesn't suck

On the first business day of every month, a thread goes live on Hacker News titled "Ask HN: Who is hiring?" Within 48 hours it accumulates 300 to 500 top-level comments. Each comment is one company describing what they're building, what the

On the first business day of every month, a thread goes live on Hacker News titled "Ask HN: Who is hiring?" Within 48 hours it accumulates 300 to 500 top-level comments. Each comment is one company describing what they're building, what they're hiring for, and — critically for anyone selling B2B tools — exactly what stack they're running and what stage they're at.

That thread is the densest buying-signal artifact published on the public internet every month. The hiring company is broadcasting their pain. They want a Kubernetes specialist because their cluster is on fire. They want a senior data engineer because their warehouse costs are exploding. They want a head of growth because revenue stalled. Every JD is a problem statement, and every problem statement is a potential conversation.

The hiring company isn't just listing a role. They're publishing a stack diagram, a team size, and a confession of what's broken — for free, every month, with timestamps.

Why this thread beats job boards

Job boards (LinkedIn, Indeed, Wellfound) list the role. They show you a polished JD written by HR or recruiting. The HN thread is different in three ways that matter for outbound.

First, the post is written by an actual employee, often the hiring manager. The voice is technical and unguarded. They'll mention that they're "moving off of Heroku because the bill got out of hand" or that "the data pipeline is held together by Airflow and prayer". That's not a JD line, that's a pain statement.

Second, the post lists the stack in plain text, usually as a bulleted "what we use" section. You can extract: cloud (AWS, GCP, etc.), infra (K8s, Terraform), data (Snowflake, Databricks), monitoring (Datadog, Honeycomb), application stack (Rails, Next.js, etc.). On a job board you'd be guessing. Here it's spelled out.

Third, the post reveals stage. "We just raised our Series B" or "we're 12 people, profitable, no funding" tells you everything about budget and decision speed. Combined with the role count, you can infer team scale within a margin.

The whole thread is structured-enough to scrape and unstructured-enough that it filters out the people who can't be bothered to actually parse it. Your competitors aren't doing this. They're buying lists and blasting.

The workflow, end to end

The whole pipeline takes under an hour for a monthly cycle if you have the tooling set up. Here's the order of operations we've seen work.

1. Scrape the thread. The thread URL is predictable: HN posts the new "Ask HN: Who is hiring?" on the first business day of each month from the user whoishiring. Pull the JSON via the Algolia HN API or the official Firebase API. You want every top-level comment with its author, timestamp, and full text.
2. Parse stack and metadata. Run each comment through a parser (regex or LLM) to extract: company name, location, remote-or-not, stack tags, role count, role types, stage (if mentioned), and any funding signal. Store this as structured rows.
3. Filter to your ICP. Apply your filters: stack contains [your trigger tools], company size in your range, geography matches, role type implies budget for what you sell. You'll typically go from 400 raw comments to 30-50 strong matches.
4. Enrich for outbound. For each matched company, you need a contact. The HN comment usually lists a contact email or links to a careers page. Use the careers-page domain to find the founder or hiring manager via enrichment. Confirm work email validity.
5. Draft the email. Three lines max. Reference the specific JD detail. Tie it to a one-sentence value prop. Ask one specific question.
6. Send and track. Drip across two weeks, not all at once. Track open, reply, meeting-booked. Re-target non-replies in the next month's thread if they post again.

The single highest-leverage step is step 5. Step 1-4 can be automated cleanly. Step 5 is where most operators fall back to a generic template and undo all the work the pipeline just did. Don't.

What the cold email actually looks like

Here's a sample email from the workflow, anonymized but otherwise verbatim. The recipient was a hiring manager at a 15-person infrastructure startup who'd posted a senior SRE role mentioning Kubernetes, Datadog, and "rapidly growing observability bill".

Subject: the Datadog bill, the SRE hire, and one question
 
Saw your Who's Hiring post — specifically the line about the
observability bill outpacing revenue. We've worked with three
teams in the same shape (K8s, Datadog, ~10-20 engineers) and
the SRE hire usually goes one of two ways.
 
Worth a 15-min call to compare notes? Or if you'd rather just
see what we did, I can send a one-pager — no demo dance.
 
— [name]

Three things to notice. The subject line cites the post directly ("the Datadog bill"). The body opens with a specific phrase from the post ("observability bill outpacing revenue"). And the ask is small — 15 minutes or a one-pager, recipient picks.

What this email does not do: pitch a product, attach a deck, ask for a meeting with three calendar links, or use the word "synergize". Every sentence is doing work. There's no preamble.

For a deeper read on the contextual-message pattern outside HN specifically, the contextual cold message walks through why quoting the source phrase verbatim is the highest-leverage move in the whole playbook.

Filters that actually narrow the noise

Most operators try to email the whole thread and burn out by month two. The right approach is to build a filter set that gets you to 30-50 matches and lets you ignore everything else.

The filter set we've seen work has four stages. Apply them in order — each one cuts the list further.

1. Stack filter. If you sell a Datadog alternative, the company has to mention Datadog (or a competitor: New Relic, Grafana Cloud, Honeycomb). If they don't mention any observability tool, they're either not your buyer or too early to know.
2. Size filter. Use the role count as a proxy. A company hiring 1-3 roles is small (5-25 people). A company hiring 8+ roles is mid-market (50+ people). Match this to your ICP.
3. Stage filter. Look for funding mentions ("just raised", "Series A", "profitable") or absence thereof. If you sell to funded startups, drop the bootstrapped ones. If you sell to bootstrappers, drop the Series B brands.
4. Geography and remote filter. Some companies only hire in specific timezones. If you sell a tool that requires onsite implementation, drop the fully-remote ones.

After all four filters, the typical funnel is something like 400 raw comments → 80 stack matches → 50 size matches → 35 stage matches → 28 final candidates. That's a manageable list to email over two weeks at 2-3 sends per day.

What HN's Who's Hiring is bad for

This thread doesn't work for every product. The clearest signal failures we've seen:

• HR tools and recruiting tech. The recipient is the hiring manager, who didn't choose the ATS and doesn't want a vendor pitch about it. Wrong target.
• Marketing and sales tools. The thread skews engineering. You'll find growth/ops roles but not in the volume you'd want.
• Late-stage enterprise sales. The companies posting are mostly under 200 people. If your ACV is $100k+ and you sell to F500, this thread is the wrong pond.
• Generic productivity tools. "We sell Notion replacement" doesn't get traction here because there's no specific JD signal that maps to your value prop.

Where it shines: infrastructure, observability, data tooling, dev productivity, security, AI/ML platforms, anything with a specific stack-replacement story.

How to fold this into a broader signal pipeline

The Who's Hiring thread is monthly. That's a problem if you need a steady flow of conversations week to week. Most operators we've seen fold this into a broader pipeline that pulls from multiple sources.

The base layer is continuous monitoring of HN — not just Who's Hiring but the front page, Show HN, Ask HN. The 5-minute HN monitoring setup walks through how to do this without paying for an SDR tool. Layered on top: Reddit signal monitoring for the niches where your buyers also post, plus LinkedIn for the people who've shifted roles.

The Who's Hiring thread becomes a monthly batch of high-quality leads that supplements the daily drip from the other sources. It's not a standalone channel — it's a force multiplier when stacked.

A two-person outbound team can realistically work the monthly thread and get 8-12 meetings booked per cycle if the ICP fit is good. That's not life-changing on its own. But added to a Reddit signal pipeline doing 5 calls a week and a LinkedIn signal pipeline doing another 5, the pipeline math gets interesting.

A note on tone

The thing that kills HN-sourced outreach faster than anything else is sounding like marketing. HN readers have an immune response to phrases like "I noticed you're hiring" or "I came across your post". Both are true and both will get your email deleted.

Use the actual phrase from their post. Use their stack words. If they wrote "we're tired of paying $X for Y", use those words. Treat the JD like a brief that the recipient wrote for you. They want to be approached by people who actually read it.

And keep it short. Three lines. The longer your email, the more you sound like the 50 other vendors who emailed the same hiring manager that morning.