Startup Ideas From Hacker News: Turn Threads Into Tests

Hacker News is a useful place to find startup ideas, but not because the front page is a market. Points are attention. Comments are context. Neither one pays your Stripe invoice.

The useful read is narrower: which threads reveal a recurring workflow, a buyer with budget, a crowded category that still needs services, or a new technical surface that makes an old pain cheaper to solve. That is where startup ideas from Hacker News become more than another saved tab.

This guide is based on SkimHQ's June 9, 2026 archive run, which captured recent Show HN and Ask HN signals, scored them, and converted the best ones into founder tests. The point is not "copy what is popular on HN." The point is to read the thread like an operator.

What the June 9 archive run found

The June 9 SkimHQ pass surfaced four useful HN signals. Each one looked different on the surface, but all four had enough public evidence to force a founder question.

• Ask HN: Why hasn't there been a real competitor to Ticketmaster yet? was captured at 123 points. SkimHQ marked it as a complaint signal, not a clean product prompt, because the thread centered on venue lock-in, fee extraction, and the difficulty of replacing an entrenched platform.
• Show HN: Gitdot was captured at 207 points. The source showed working signups, org creation, private and public repos, and GitHub import or mirror flows. SkimHQ read that as evidence that self-hosted Git alternatives are becoming practical enough for cost and seat-optimization wedges.
• Show HN: I Derived a Pancake was captured at 331 points. The obvious read is "people like obsessive food science." The founder read is that small food operators may pay for repeatable formulation help when recipe work moves from kitchen experiment to production process.
• Show HN: Performative-UI was captured at 885 points. SkimHQ read the attention as interest in interface patterns that shape perceived speed and trust, then tested whether that could become a merchant-facing conversion toolkit rather than a developer curiosity.

None of those signals is enough by itself. HN can over-reward novelty, technical cleverness, and jokes. The value comes from the second pass: evidence type, payer, wedge, week-one action, and kill signal.

The Hacker News signal stack

How SkimHQ turns an HN thread into a founder test

SkimHQ does not treat every front-page thread as an idea. The archive pipeline first classifies the evidence type. A complaint thread is different from a working demo. A working demo is different from a procurement clue. A comment pile full of jokes is different from one where operators describe exactly where the workflow breaks.

SkimHQ does not treat every front-page thread as an idea.

After that, the system asks four blunt questions.

1. What is the actual evidence?

The Ticketmaster thread was not proof that a new ticketing platform will beat the incumbent. It was proof that people understand the pain and keep asking why the market has not opened. That is a complaint signal. SkimHQ marked it as crowded and pushed the idea away from "build a competitor" toward a service-led migration wedge for smaller venues.

2. Who pays?

"HN readers" almost never pay. The buyer has to sit outside the thread. For Gitdot, the payer is not the developer who likes Rust. It is an engineering manager or IT owner staring at repository access, inactive seats, and budget pressure. For Performative-UI, the payer is not the React developer applauding a clever library. It is the merchant who wants better checkout behavior without custom frontend work.

3. What is the narrow wedge?

Good HN research gets smaller after the first pass. Gitdot becomes a seat audit or mirror strategy, not "replace GitHub." Pancake chemistry becomes recipe-formulation help for small producers, not a consumer recipe app. Performative-UI becomes three conversion patterns packaged for one commerce platform, not a universal design system.

4. What kills the idea quickly?

The best output from a signal scan is often a kill signal. Email 20 venues and get fewer than three real switching conversations? The ticketing wedge is weak. Offer three free recipe optimizations and nobody submits a real recipe? The pancake enthusiasm was entertainment, not buyer urgency. HN can start the test. It cannot be the test.

Four founder-grade reads from the June 9 run

What weak HN research gets wrong

Weak HN research confuses audience fit with market fit. A technical audience liking a technical demo is useful, but it does not mean a budget owner has pain. It also does not mean the product should target HN readers.

The other common mistake is copying the artifact. If a source project goes viral, most founders ask whether they can build a similar project. That is usually the worst question. The better question is what the source project makes easier, what adjacent work it reveals, and which buyer already has a reason to care.

• Do not rank by points alone. A 120-point complaint from operators can be stronger than a 900-point novelty if the payer is clearer.
• Do not skip the buyer. If the payer is "developers maybe," the idea is still foggy.
• Do not build before outreach. The first test should usually be a teardown, audit, service offer, or paid pilot.
• Do not ignore category risk. Crowded, regulated, or contract-heavy markets need a service wedge before they deserve product work.

A 45-minute Hacker News idea scan

Use this when you want a repeatable scan instead of an hour of interesting procrastination.

Use this when you want a repeatable scan instead of an hour of interesting procrastination.

Minute 0-10: collect only recent operator threads

Start with Ask HN, Show HN, and high-comment threads from the last week. Skip pure launch announcements unless the comments expose a workflow. Log title, URL, points, comment count, source type, and the exact sentence that points to pain.

Minute 10-20: classify the evidence type

Mark each source as complaint, working demo, workflow clue, budget clue, or weak signal. This keeps you honest. A cool demo is not a complaint. A complaint is not proof that buyers will switch. Each type needs a different next step.

Minute 20-32: name the payer and their current workaround

Write one sentence: "The buyer is [specific role] who currently handles [workflow] by [current workaround]." If you cannot name the role and workaround, the thread is not ready. Save it, but do not build from it.

Minute 32-45: define the smallest paid test

Reduce the idea to one buyer action. A venue fee audit. A repository access review. A recipe optimization. A conversion widget install. Then write a kill signal before you start. The output should make action easier, not add another idea to your backlog.

Where this fits with SkimHQ

Hacker News is one source surface. It gets better when compared with other evidence: Reddit complaints, crowdfunding commitment, search intent, public records, job posts, and competitor gaps. That is why this post sits beside the Reddit signal-scoring guide, the Kickstarter demand guide, and the buyer-proof checklist.

The operating habit is the same across sources. Collect the signal, classify the evidence, name the buyer, define the smallest test, and kill weak ideas quickly. That is the difference between a founder intelligence workflow and another tab folder called "ideas."