A buying-intent score should explain why a conversation deserves attention; it should not hide judgment behind a number. This rubric uses five observable dimensions—fit, action, urgency, specificity, and access—scored from zero to two. The total helps order the queue, while counter-signals can still override it. The examples are representative and illustrate the method rather than reported conversion outcomes.
Score evidence, not optimism
Award points only for details visible in the conversation. Do not infer budget, authority, or urgency because the account looks promising.
One strong phrase cannot carry the score
A request for alternatives may show action, but weak fit, no context, or an inaccessible community can still make the thread a poor opportunity.
Counter-signals protect the queue
A stale thread, solved request, student project, unsupported market, or prohibition on vendor participation can outweigh a high raw total.
Thresholds should reflect capacity
A solo founder may review only the highest band. A research team can keep lower-scoring threads for language mining without turning them into outreach.
The five dimensions of an auditable intent score
Give zero when evidence is absent, one when it is partial, and two when it is explicit. A ten-point total is a prioritization aid, not a promise of conversion.
Fit
Audience, workflow, geography, and product capability match the problem described.
Implication: A score of two requires a problem the product can solve today without stretching its promise.
Action
The buyer is comparing, replacing, shortlisting, requesting recommendations, or planning a migration.
Implication: Category discussion alone scores zero; an explicit decision behavior scores two.
Urgency
A renewal, launch, deadline, active failure, or near-term implementation window is visible.
Implication: Urgency should come from stated timing, not the emotional tone of the post.
Specificity
The thread names constraints, required outcomes, failed approaches, or evaluation criteria.
Implication: Specificity makes a useful response possible and reduces guesswork.
Access
The source is current, public, context-rich, and permits transparent participation.
Implication: A commercially strong thread can still score zero on access and become research-only.
9/10 representative: active shortlist
A SaaS founder wants a lighter way to find recommendation threads, names current-tool noise, asks for alternatives, and needs a decision this week. The community permits vendor replies.
Why it matters: Fit 2, action 2, urgency 2, specificity 2, access 1. Review quickly, but still lead with a useful comparison rather than a pitch.
7/10 representative: specific pain without timing
A small team describes a precise monitoring problem and asks how others solve it, but does not indicate a purchase or deadline.
Why it matters: Fit and specificity are strong; action and urgency are partial. A clarifying answer or research classification is more appropriate than sales pressure.
5/10 representative: explicit alternative request, weak fit
An enterprise team asks for a replacement but requires regulated-media coverage and reporting beyond the product's scope.
Why it matters: Action and urgency cannot compensate for low fit. Reject the opportunity instead of optimizing the score around attractive language.
4/10 representative: useful problem research
A founder explains a recurring workflow frustration in detail but is building an internal solution and is not considering software.
Why it matters: The specificity is valuable for content and product research, while action and urgency remain absent.
Raw 8/10, overridden: participation is prohibited
The problem, action, urgency, and specificity are clear, but the private community prohibits solicitation and does not permit vendor participation.
Why it matters: Access overrides the raw score. Do not contact the buyer outside the community or treat public-looking intent as consent.
Store the dimension scores with the total
A total of seven can mean several different things. Preserve the five components so reviewers can see whether fit, urgency, or access created the result.
Set separate thresholds for reply and research
For example, reserve the highest band for immediate human review and keep specific but lower-motion threads in a research queue. Tune the bands to capacity rather than copying a universal cutoff.
Audit overrides every week
Review threads whose raw score was high but the final action was reject. Repeated override reasons often expose a missing dimension or an overly generous rule.
Test the rubric on one source first
Start with a source such as Reddit where thread context and community rules are visible, then adapt the access and specificity rules before adding other platforms.
Prioritize public buying intent without hiding judgment behind a single number.
ReplyRadar brings fit, conversation context, and intent clues into a reviewable queue so founders can decide what deserves action.
What is a good buying-intent score?
There is no universal number. A useful threshold depends on the source, product, team capacity, and action. Calibrate it against human decisions and preserve override reasons.
Can AI determine public buying intent automatically?
AI can help classify visible evidence and prioritize a queue, but product fit, community permission, ambiguity, and the final reply decision still benefit from human review.
Should negative sentiment increase an intent score?
Only when it contributes observable movement or urgency. Strong emotion by itself is a complaint signal, not proof that the buyer is evaluating a replacement.