Guide
Buying intent signals, explained
Buying intent is not a keyword score. It is a public statement of a problem, a constraint, and a decision window. If you cannot point to those three things in the post, you are looking at a topic, not a buyer.
A working definition
A buying intent signal has three parts: a problem the author wants solved, a constraint that makes the current option insufficient, and a reason the decision is happening now. Remove any one and the signal weakens.
| Part | Question | Example |
|---|---|---|
| Problem | What is broken or missing? | “Our Postgres connection pooler keeps dropping.” |
| Constraint | Why is the current option not enough? | “We need self-hosting for compliance.” |
| Window | Why now? | “We launch in three weeks.” |
The author does not need to say “I am buying.” The combination is enough. A post that names a problem but no constraint is a complaint; one that names a constraint but no window is research.
The five signals worth scoring
Score each 0–20. The total is a prioritization tool, not a prophecy. Keep the reasons, not just the number: a score without a reason cannot be reviewed, improved, or explained to a teammate.
| Signal | 20 | 0 |
|---|---|---|
| The ask | An explicit request for a recommendation | No question at all |
| Specificity | Names the stack, budget, or constraint | “Any tools?” |
| Recency | Hours to a few days | Months old |
| Context | Founder, team, or budget holder | Student, researcher, or competitor |
| Stage | Comparing, migrating, or buying | Announcing a launch |
The false positives that waste a week
- Listicles: “10 best X tools” contains every keyword and no buyer.
- Launch posts: “Show HN: I built X” is a competitor or a peer, not a buyer.
- Students and researchers: real questions, wrong context.
- Competitor research: “How does X compare to Y?” may be a buyer or a rival. Read the account.
- Vents without an ask: a rant can be a signal, but only if the author asks for a way forward.
A filter that drops 90% of matches is working. The goal is not coverage; it is a short list you will actually answer.
A rubric you can defend
A simple rubric is better than a model you cannot explain. Start with the five signals above, 0–20 each, and a threshold of 60.
| Score | Meaning | Action |
|---|---|---|
| 80–100 | Active problem, clear constraint, near-term decision | Reply today |
| 60–79 | Real question, weaker urgency | Reply this week |
| 40–59 | Topical, not buying | Save the search, not the thread |
| 0–39 | Noise | Drop |
Review the threshold after a month. If you are replying to threads that go nowhere, raise it. If you are missing good conversations, lower it. The number is a lever, not a law.
Where to capture signals
The best source is where your buyer already asks questions. For developer tools that is usually Reddit, Hacker News, GitHub issues, and the support forums of the tool you replace.
- Reddit: problem-specific subreddits, sorted by new.
- Hacker News: comments on Ask HN and migration threads.
- GitHub: issues that say “alternative”, “recommendation”, or “how do you”.
- Review sites: the “cons” section of a competitor’s reviews.
- Search: queries that include “alternative”, “recommendation”, or “worth it”.
Turn signals into a habit
- Pick two sources and three query patterns.
- Check them on a fixed cadence, not continuously.
- Score every candidate and write one sentence of reasoning.
- Reply to the top few, from your own account, with disclosure.
- Log the outcome and prune the queries that produce noise.
The habit is the system. A tool that finds a thousand posts you never read is worse than a search you check for ten minutes a day.
How LaunchEngine scores
LaunchEngine uses a versioned classifier to score each post 1–100 on buying intent, keeps the reasons, and defaults to a threshold of 60. It scans Reddit, Hacker News, and the open web, and drafts a reply you review and post.
The score is a prioritization tool. The draft is a starting point. The decision to reply is yours.
Questions
- What is a buying intent signal?
- A public statement where someone describes a problem they want solved, a constraint that makes their current option insufficient, and a reason the decision is happening now. A keyword match without those three parts is a topic, not a buyer.
- How do you score buying intent?
- Score five things: whether there is an explicit ask, how specific the constraint is, how recent the post is, who the author is, and what stage they are in. A 0–20 score for each, with a threshold around 60, is enough to prioritize a short list.
- Why do keyword alerts produce so many false positives?
- Keywords match topics, not decisions. Listicles, launch posts, and competitor research contain the same words as a buyer. A scoring step that reads the post and keeps the reason is what separates the two.
- What should I do with a high-intent signal?
- Read the whole thread, answer the question without pitching, disclose your affiliation if you mention your product, and post from your own account. If the thread is already resolved, stop.
Find your first conversation
LaunchEngine scans Reddit, Hacker News and the open web, scores each post for buying intent, and drafts a reply you review and post. Start with 100 credits, one project and one scan a day.
Keep reading
5 min read
How to find people asking for your product
A practical workflow for finding public buying-intent conversations on Reddit, Hacker News and the open web, and deciding which ones are worth answering.
4 min read
Reddit lead generation for SaaS
How to find Reddit threads where people are asking for a product like yours, reply without getting banned, and measure conversations instead of clicks.