Review sooner
Strong score + specific signals
The available text shows stronger fit and intent evidence relative to other matches.
Read the full source, verify the need, and decide whether research or a useful response belongs.
Feature · qualification
Stop scanning every mention. AI Lead Scoring brings recommendation requests, comparisons, switching language, and urgency to the top of your queue, so you can decide what deserves attention first.
Written by Michael Yin
IntentLoom turns the context available to the scorer into a 1–100 priority. The number is useful because the rubric stays inspectable—not because it predicts who will buy.
Product fit
30%
Does the need align with what the project can actually solve?
Intent type
20%
Is this a recommendation, comparison, switching, or purchase-stage conversation?
Pain
15%
Is a concrete problem affecting the author now?
Authority
15%
Does the author appear close to the decision or only adjacent to it?
Urgency
10%
Is there a deadline, active evaluation, or immediate reason to act?
Budget
5%
Does the available text reveal price sensitivity or spending context?
Community context
5%
Does the subreddit and discussion context support a relevant response?
Lead Criteria
A general scoring rubric can evaluate relevance, intent, pain, authority, urgency, budget, and community context. Your business may still have clear requirements or exclusions that those dimensions cannot capture on their own.
Lead Criteria lets you describe those rules in plain language. IntentLoom considers them while scoring matched posts and comments. Preferences can shape the score, while a clear exclusion supported by the Reddit content can keep a conversation out of the lead workflow.
How to read the result
The structured result can include a score, intent stage, confidence, reason, and positive, negative, or disqualifying signals. Those details explain why the item moved—not what you must do.
Review sooner
The available text shows stronger fit and intent evidence relative to other matches.
Read the full source, verify the need, and decide whether research or a useful response belongs.
Do not force it
The author may be shopping, but not for what your product provides.
Treat fit as the veto. Learn from the conversation or skip it.
Inspect manually
Sparse, ambiguous, sarcastic, or conflicting context makes the ordering less reliable.
Read it yourself or keep it out of the active response queue.
For a post, the current scorer works from the subreddit, title, and prepared summary. For a comment, it uses the subreddit, parent post title, and comment body. It does not silently know the author, their budget, or the complete external story.
These controls reduce overconfidence; they do not eliminate false positives. Human review remains the final qualification step.
What stays in your hands
IntentLoom reduces repetitive discovery and review work. You still decide what the conversation means, whether a response belongs, and what becomes public.
It helps surface context such as recommendation requests, comparisons, switching language, frustration with a current tool, and urgency. These are signals to review, not a promise of conversion.
No. The score helps order your attention, but you should read the surrounding conversation and decide whether a response is relevant and appropriate.
Use the context to decide whether to research, draft a reply, discard the match, or route the qualified opportunity into your notification workflow.
The current rubric considers product fit, intent type, pain, authority, urgency, budget, and community context. Confidence can reduce the result, and weak product fit limits how high a conversation should rank.
No. It means the available context has stronger qualification signals relative to other matches. The author may still be researching, may not fit your product, or may not welcome a vendor response.
Yes. Lead Criteria lets you describe preferences and clear exclusions for each project in plain language. IntentLoom considers them when scoring matched posts and comments. A clear exclusion is applied only when the Reddit content provides supporting evidence. Human review remains necessary for the leads that continue through the workflow.
Find relevant Reddit conversations without treating every match as a lead.
Qualify matched threads for real buying intent before you reply.
Draft a relevant reply while keeping the final public action human.
Keep the final action human