Most outbound fails because operators reply to everything. They see a keyword match, draft a response, and send — without classifying whether the match is actually a buying signal worth a reply. The result: low reply rates, burned hours, and the channel gets blamed when the targeting was the actual problem.

The 1–10 buying intent score is the framework that fixes this. It's a simple methodology you can run in your head (or, when you scale, automate via tools like repco's intent classification) that cuts your reply queue to the 30% of matches actually worth replying to. This post covers the scoring rubric, the four intent types that determine action, examples at each score, and how to apply it across Reddit, LinkedIn, and any other public signal channel.

Key takeaways

• The 1–10 framework classifies every signal on intent strength; only signals scoring 6+ are worth replying to.

• Four intent types separate signals into action categories: direct (highest), competitive, problem, engagement (lowest).

• ~30% of keyword matches typically score 6+; 50% score 4–5 (skip); 20% below 4 (noise).

• Recency multiplies score — a 9 from 30 minutes ago is a different signal than a 9 from 30 hours ago.

• The framework works manually and is the same one Claude Sonnet 4.6 uses inside repco for automated classification.

What is the 1–10 buying intent score?

The 1–10 buying intent score is a quick-classification framework that rates every signal on how strong the buying intent is and how urgent the reply should be. The score combines two dimensions — intent strength (how clearly the post indicates someone wants to buy) and recency (how fresh the signal is) — into one number that drives the action decision.

The full rubric:

In practice ~30% of keyword matches score 6+, ~50% score 4–5, ~20% score below 4. The bottom 70% never gets a reply. The output you actually act on is the top 30% — typically 15–30 high-quality signals/day for a properly configured watchlist.

The framework works because it forces a decision before drafting. Most operators write the DM first and decide whether to send it second, which is backwards. Score first, decide whether to draft, then write only if the score justifies the time.

The 4 intent types that determine action

Intent score is one dimension; intent type is the other. Same score can mean different actions depending on what kind of signal it is. Four types cover most B2B buying signals you'll see:

Direct intent

Someone explicitly asks for a product like yours. Names your category, names a constraint, or asks for a recommendation in plain language.

Examples:

• "Anyone know a good CRM for solo founders that isn't HubSpot?" (Score 10)

• "Looking for an alternative to Apollo — same database, less spam." (Score 9)

• "What are people using for cold email warmup these days?" (Score 8–9)

Action: DM within the hour. Use Template 1 or 2 from the Reddit DM templates family.

Competitive intent

Someone names a competitor and expresses frustration, criticism, or a desire to leave. They haven't asked for alternatives explicitly, but the intent to switch is clear.

Examples:

• "Phantombuster pricing is a joke now." (Score 8–9)

• "HubSpot's onboarding is a nightmare and I'm questioning the contract." (Score 7–8)

• "Apollo lists keep getting worse — starting to wonder if it's worth the renewal." (Score 8)

Action: Same-day reply. Use Template 3 or 4. Don't pile on the competitor by name beyond what they wrote — sounds petty.

Problem intent

Someone describes a problem your product solves but doesn't name a tool or category. Lower targeting confidence than direct or competitive intent, but high volume.

Examples:

• "Spent 6 months trying to fill a pipeline as a solo founder and nothing's working." (Score 7)

• "Reply rates dropped 70% this year, what am I doing wrong?" (Score 7–8)

• "Domain got burned again, third one this year." (Score 6–7)

Action: Same-day reply if you can add real diagnostic value. Use Template 5 or 6. Lead with the diagnosis, not the pitch.

Engagement intent

Someone is engaging with topics adjacent to your category but isn't actively in market. Long-game pipeline building, not immediate-conversion play.

Examples:

• "Curious about LinkedIn outreach in 2026 — anyone running it at scale?" (Score 5–6)

• "Reading this thread about Reddit for B2B — anyone using it seriously?" (Score 5)

• "Why is everyone leaving cold email?" (Score 5–6)

Action: Reply within 48 hours only if you can add real value. Use Template 7 or 8. Plant a flag, don't pitch.

How to score in your head: the 30-second rubric

For every signal, run through five questions in order. The first "yes" determines the score:

1. Did they explicitly ask for a product like yours? → Score 9–10

2. Did they name a competitor with frustration or switch intent? → Score 7–9

3. Did they describe a problem your product solves, no tool named? → Score 6–8

4. Is it adjacent to your category but no clear ask? → Score 4–6

5. Off-topic or vague mention? → Score 1–3

Then apply the recency adjustment:

• Posted within 6 hours: +1 to score

• Posted within 24 hours: no adjustment

• Posted within 48 hours: -1 to score

• Older than 48 hours: skip regardless of score

This takes about 30 seconds per match in your head. Over a session of 50 keyword matches, you score them down to ~15 worth replying to in roughly 25 minutes.

Why most operators score too high

The most common scoring failure is grade inflation — calling something a 9 when it's actually a 6. Three patterns drive it:

1. Wishful thinking on targeting. Operators want their watchlist to be valuable, so they convince themselves a tangential mention is a strong signal. Cold-eye it: would a stranger reading this post agree the person is asking for your product specifically?

2. Confusing recency with intent. A post from 30 minutes ago feels urgent, but if the intent is weak (vague rant, no ask), recency doesn't make it a 9. Recency is a multiplier on real intent, not a substitute for it.

3. Mistaking adjacent fit for direct intent. Your product could help with their problem, but they're not asking for your category. That's a 5–6, not an 8. The action is different (long-game vs. immediate-DM) and the messaging has to match.

The fix: be honest. Score against the rubric, not against your hopes. The 30-second pause to score reduces wasted DMs by 60–70% and protects your account health from spamming low-intent threads.

How recency multiplies the score

Recency matters because thread engagement decays fast. Reddit's own data shows comment engagement collapses 90%+ after the first 6 hours. LinkedIn posts have a longer half-life (24–48 hours) but the same dynamic applies. A reply that lands after the conversation has moved on doesn't get read — doesn't matter how good the message is.

The practical implication: a 9-rated direct ask from 30 minutes ago demands an immediate reply (within the hour). The same 9-rated post from 30 hours ago is at most an 8 — still worth replying, but the urgency is gone and the reply will get less traction. From 48 hours ago, even a perfect-intent post drops to a 6–7 and is often not worth the time.

This is why repco watches Reddit every 15 minutes and LinkedIn every 2–4 hours — the cadence is platform-aware so high-intent recent signals don't go stale before they reach the inbox.

Applying the framework across channels

The 1–10 framework works across any public signal channel — Reddit, LinkedIn, niche communities (Slack, Discord, Indie Hackers), even Twitter/X. Three channel-specific notes:

Reddit. Highest signal density per match because the platform structures conversations around questions and recommendations. The 4 phrase patterns catch ~80% of buying intent. Score breakdown skews higher than other channels because Reddit's question-asking culture surfaces direct intent more often.

LinkedIn. Signal often arrives as comments under tool reviews, posts about industry events, or rants about current vendors. Direct asks (Score 9–10) are rarer; competitive intent (Score 7–9) is more common. Recency multiplier matters less because LinkedIn posts have longer half-lives.

Communities (Slack/Discord/Indie Hackers). Highest noise floor because the communities mix social chat with commercial intent. Score harder — a casual mention in a community is rarely a 6+ unless it's an explicit ask.

The complete 2026 outbound guide covers how to wire each channel into a multi-channel pipeline once you've validated the framework on one.

Frequently asked questions

Why 1–10 instead of just "reply or skip"?

A binary classification (reply / skip) loses too much information. The 1–10 scale lets you prioritize within the "reply" set (DM the 9s within the hour, the 7s same-day) and within the "skip" set (5s become reply candidates if you have time, 3s are firm skips). Granularity matches the variance in signal quality you actually see in practice.

Can I automate this scoring?

Yes — Claude Sonnet 4.6 inside repco runs the same framework on every match, returns intent_strength (1–10), intent_type (direct/competitive/problem/engagement), and a one-sentence reasoning trace. The automation is the same logic running 24/7 across more channels than you can watch by hand. We covered the two-tier detection stack in detail.

What if I get the score wrong?

The framework is a fast first-pass filter, not an oracle. You'll mis-score sometimes — calling something a 7 that's really a 5, or skipping a 6 that turned out to be a meeting. The point is reducing the median error: most operators reply to a 5 thinking it's a 9, which wastes time and account credibility. Even a noisy 1–10 framework is better than no framework.

Does the framework work for cold email signals?

No — cold email signals are different. There's no public intent signal because the prospect didn't ask. Cold email targeting is fit-based (does this account look like our ICP), not intent-based. The 1–10 framework is designed for public-channel intent monitoring (Reddit, LinkedIn, communities). For why cold email economics broke, the issue isn't scoring — it's that broadcast outbound stopped working.

Related reading

• For how to operate against intent signals once you have scored them, see the signal-based selling 2026 playbook - 5-step motion calibrated for solo founders.

• For the 7 main intent data sources compared (Bombora, 6sense, ZoomInfo, G2, Common Room, repco), see intent data for B2B sales.

• For the 5 AI use cases that work in prospecting (including signal classification and scoring), see AI for sales prospecting in 2026.

Bottom line

The 1–10 buying intent score is a 30-second filter that cuts your reply queue from "every keyword match" to "the 30% worth replying to." Combined with the four intent types (direct, competitive, problem, engagement), it tells you both whether to reply and how to reply.

Most outbound fails because operators skip the scoring step. Adding it back — manually or via intent-driven automation — produces 5–10x reply rates compared to replying to everything. Same effort, fundamentally different outcome.

Apply it on Reddit (the manual playbook) before scaling to LinkedIn (the discipline rules). The complete outbound guide covers where this fits in the bigger picture.

About the author

Kamil is the founder of repco.ai — the AI sales rep that finds buyers publicly asking for products like yours on Reddit and LinkedIn. 15 years across marketing and sales, building and running companies in industrial, IT, investments, and real estate. Serial founder; building repco from the gap he kept hitting himself — outbound channels that work for solo founders and small teams, not enterprise sales orgs. Designed and tested the 1–10 framework manually for 18 months before encoding it into repco's automated classification.