How Do You Score Prospects by LinkedIn Engagement?

Flux's score_connections ranks every person in your engagement graph by warmth — a composite of how recently they engaged, how often they come back, how senior they are, and whether they commented or just liked. Combined with constellation mapping and an outreach queue, it turns scattered LinkedIn activity into a prioritized list of people worth reaching out to, ordered by who is most likely to respond.

What you'll learn

How score_connections weights engagement signals into a warmth score
Setting up segments to define your target audience slices
The full workflow: target accounts, constellation mapping, scoring, outreach queue, outcome logging
What makes someone "warm" versus "lukewarm" — interpreting the scores
Closing the feedback loop with log_outcome

How does warmth scoring work?

Edge

score_connections evaluates every person who has engaged with your content and assigns a warmth score based on four weighted factors:

Recency. Someone who commented on your post yesterday scores higher than someone who liked a post six months ago. Engagement decays — a warm lead today is a cold one in three weeks if there is no follow-up.

Frequency. A person who engaged with five of your last ten posts is a stronger signal than someone who appeared once. Across Flux's dataset of 25,150 posts from 66 profiles, repeat engagers convert to conversations at a meaningfully higher rate than one-time reactors.

Engagement type. Comments carry more weight than reactions. Someone who wrote a thoughtful reply invested real attention; someone who tapped "like" may have been scrolling. The scoring reflects this — a single comment can outweigh several likes.

Seniority and title fit. A Director of Sales commenting on your content about pipeline tools is a higher-value signal than a student liking it. The scoring factors in title-level seniority so that decision-makers surface above the noise.

Run it against your engagement graph to get a ranked list:

score_connections(profile_url: "https://www.linkedin.com/in/yourprofile")

The output is a scored roster — each person with their warmth score, company, title, and the engagement signals that drove the score. You are looking for people who score high on multiple factors simultaneously: recent, frequent, senior, and commenting.

How do I define who counts as a prospect?

Scoring your entire engagement graph is useful, but most GTM workflows start with a target list. Flux segments let you define audience slices using engagement-based predicates — "people who commented on at least 2 posts in the last 30 days" or "VPs and above from companies with 500+ employees."

Create a segment that matches your ideal customer profile:

create_segment(name: "Enterprise VPs - Recent Commenters", predicates: {...})

Segments are reusable. Once defined, you can score only the people in a segment rather than your full graph, which keeps the results focused on accounts you actually want to pursue. You can also layer segments — one for seniority, one for engagement frequency — and intersect them.

Think of segments as the "who matters" filter that sits upstream of scoring. Scoring tells you how warm someone is; segments tell you whether they are worth scoring at all.

What is the full scoring workflow?

The end-to-end flow moves from broad mapping to a specific, prioritized action list in five steps:

1. Define target accounts. Start by identifying the companies you want to reach. You can do this manually or by pulling companies from engagement data — get_post_engagers_by_company on your recent posts reveals which companies are already paying attention.

2. Map the constellation. map_constellation discovers connection chains between you and people at target accounts. It finds mutual engagers, alumni connections, and overlapping audiences — the relationship fabric that makes outreach warm instead of cold.

map_constellation(target_companies: ["Acme Corp", "Globex Inc"])

3. Score connections. Run score_connections against the constellation to assign warmth scores. The people with the highest scores are the ones with the strongest existing relationship signal — they have engaged recently, repeatedly, and substantively.

4. Pull the outreach queue. get_outreach_queue returns a prioritized list of people to contact, ordered by warmth score. This is your daily or weekly action list — the specific humans worth reaching out to right now.

get_outreach_queue()

5. Log outcomes. After you reach out, log_outcome records what happened — replied, booked a call, no response. This feedback loop sharpens future scoring by teaching the system which signals actually predicted a response.

log_outcome(person: "...", outcome: "replied")

What do the scores actually mean?

Warmth scores are relative to your own engagement graph, not an absolute benchmark. A score that is "high" for one profile may not be comparable to another, because the scoring is calibrated against your audience.

High warmth means the person has engaged recently, multiple times, with comments, and holds a title that suggests decision-making authority. These are people who already know your thinking and have demonstrated interest. A direct message referencing a specific post they commented on has a strong chance of getting a reply.

Medium warmth typically means one or two of the four factors are strong but the others are weak — someone senior who engaged once two months ago, or a frequent reactor who is not in a decision-making role. These are worth nurturing with content rather than cold-pitching.

Low warmth means minimal engagement signal. The person may be in your network but has not interacted with your content in a meaningful way. Scoring surfaces them for completeness, but they belong at the bottom of your outreach priority.

The practical threshold: focus your outbound time on the top tier. If you have 15 high-warmth prospects and 200 low-warmth ones, the 15 will generate more pipeline than all 200 combined.

How does this connect to outreach tools?

The outreach queue is designed to feed directly into whatever outreach system you use. get_outreach_queue gives you structured data — name, company, title, warmth score, and the engagement evidence behind it — that you can use to personalize messages. Pair it with get_outreach_path to see the specific connection chain (who engaged, when, on which posts) so your message can reference something real instead of opening cold.

For teams using Flux's MCP integration with Claude or Cursor, the entire workflow — from constellation mapping through scoring to drafting personalized outreach — can run in a single conversational session.

FAQ

Does scoring cost credits? Yes. score_connections and map_constellation consume credits from your monthly Edge allowance of 200. A typical scoring run across a focused target list uses a small fraction of that budget.

Can I score people who have not engaged with my content? No. Warmth scoring requires engagement data — it measures how someone has interacted with your posts. If a person has never engaged, there is no signal to score. Use get_profile_summary for a prospect brief on people outside your engagement graph.

How often should I re-score? Weekly works well for active posters. Engagement data refreshes daily on paid plans, so scores stay current. Re-scoring weekly catches new engagement patterns without over-rotating on single interactions.

Does a comment always beat a like in scoring? Comments are weighted more heavily than reactions, but frequency and recency can compensate. Someone who has liked your last eight posts scores higher than someone who left one comment three months ago.

Can I export the outreach queue? Yes. The outreach queue returns structured data that can be exported or piped into other tools through the MCP integration.

On this page