The Compound Signal Score: How to Score LinkedIn Signals Before You Send a DM

Most founders send DMs blind. They see a profile view or a comment, then reach out with a pitch. The response rate on those messages is single-digit for a reason. The recipient has no context, no relationship history, and no signal warmth. You're walking up to someone at a conference and asking them to buy before you've exchanged names.

The founders I work with who close deals from LinkedIn do it differently. They score the relationship before they send a single message. They know which connections are warm, which are cold, and which sit one interaction away from converting. They don't guess. They quantify.

The Compound Signal Score is a three-dimensional system for measuring LinkedIn relationship strength. It weights engagement depth, recency, and relevance into a single number. That number tells you exactly when a connection is ready for outreach. No intuition required.

What Is the Compound Signal Score?

The Compound Signal Score is a 0-to-10 number that measures how warm a LinkedIn relationship is right now. It blends three weighted inputs: engagement depth, signal recency, and ICP relevance. A score of 7 or higher opens the outreach window; below 5, you build signal instead of sending a DM.

Why Most LinkedIn Outreach Fails Before You Type a Word

The standard founder playbook goes like this. Connect with someone. Wait a week. Send a DM that mentions something vague about their profile. Hope for a reply. Timing decides whether that DM gets a reply, not the message. They're messaging someone whose relationship score is zero.

Think about how real business relationships form. You meet someone at an event. You see them again a month later. They comment on an article you wrote. You swap a few emails. Only after several touchpoints does a business conversation happen naturally. No one walks up to a stranger and pitches in the first 30 seconds. Yet that's exactly what most LinkedIn outreach does, and I did it for months before I saw the pattern.

The Compound Signal Score replaces intuition with structure. It asks three questions: how deeply has this person engaged, how recently did that happen, and how relevant is their world to yours. The answers produce a number from 0 to 10. Anything below 5 is cold. Don't DM; build score. Anything above 7 is warm. That's your outreach window.

The Three Dimensions of the Compound Signal Score

The score has three components. Each is measured on a 0-to-10 scale and weighted differently. Not all signals are equal. A thoughtful comment matters more than a profile view. A recent interaction matters more than one from six months ago. This is the same logic behind the Koka Signal Model, applied to a single relationship instead of a whole feed.

Dimension 1: Engagement Depth (Weight: 50%)

Engagement depth measures how meaningful each interaction has been. A profile view is a 1. A comment on your post is a 3. A DM where they reached out first is a 6. A call or meeting is an 8. The scale rewards two-way, substantive contact over passive behavior.

Engagement has tiers, and most founders flatten them. They treat every notification the same. A profile view signals curiosity. A comment signals engagement. A DM signals motivation. Scoring them identically is the fastest way to mistime your outreach.

Dimension 2: Recency (Weight: 30%)

Signals decay. A comment from last week is worth more than a comment from six months ago. The recency dimension applies a time-decay function to every interaction. Engagement within the past 7 days earns full value. 8 to 30 days earns 70 percent. 31 to 90 days earns 40 percent. Beyond 90 days earns 10 percent, unless the depth was an 8 or higher.

This is where I watch founder outreach break. They see a comment from three months ago and treat it as warm. It isn't. The signal has decayed, the connection has moved on, and you're starting from scratch. When a high-value connection goes cold, rebuild signal before you reach out. Comment on their posts. Share their work. Then DM.

Dimension 3: Relevance (Weight: 20%)

Relevance measures how aligned a connection is with your business. Someone in your industry scores higher than someone outside it. Someone at a target account scores higher than someone at a random company. Someone whose title maps to your ICP scores higher than a role that will never buy.

Relevance carries the lowest weight because a deep, recent relationship with an adjacent contact still beats a shallow one with a perfect-fit prospect. But relevance tells you where to spend your engagement energy. Of your 2,000 connections, my program clients score the 100 to 200 people worth actively building depth with.

The Account-Level View: Signals That Stack Across a Buying Group

Individual scores are useful. Account-level patterns change how you work a territory. When several people from the same company engage with your content, that isn't coincidence. That's a buying group doing its research in public. B2B deals rarely close through one person; they close through a committee. Your scoring should see the committee, not just the contact.

Here is how it plays out in my own pipeline. A VP of Sales likes a post on pipeline strategy. On her own, that's a weak signal. A week later, a sales director at the same company comments on a post about discovery. Two people, one company, ten days. Three days after that, the VP shares an article on founder-sourced revenue. Now the pattern is a buying signal, and the account has moved to the top of my list.

Two people don't research a vendor by accident. When they surface across separate content pieces close together, treat the account as warm even if no single person is. I've watched this pattern predict deals that no individual score would have flagged. The account view is where compound signal scoring earns its name.

How to Score Your Network in 30 Minutes a Week

Scoring 2,000 connections by hand isn't realistic. You don't need to. Score the 100 to 200 high-relevance connections that make up your pipeline universe.

Start by exporting your connections and filtering for ICP relevance: title, company size, industry. That produces 100 to 200 names, not 2,000. Then pull three data points per name: the last engagement type, how recent it was, and an ICP relevance score. Most of this lives in your notifications, DMs, and memory. I keep mine to three columns, and filling them for 200 names takes about 30 minutes.

The output is a ranked list. Sort by score, descending. The top 20 are your warm pipeline. The next 50 are your nurture list. The rest sit on a build list. Each list gets different treatment. Warm pipeline gets personalized DMs this week. The nurture list gets targeted engagement. The build list gets passive observation until a signal moves them up.

The 7-Day Signal Building Protocol

When a connection sits between 5 and 6, you're one engagement away from the outreach window. Don't DM yet. Run this protocol instead.

After 7 days, re-score. If engagement depth rose and recency reset within 7 days, the score clears 7. Now you DM. The message lands differently because you're not a stranger. You're the person who engaged thoughtfully for a week. That gap is the difference between single-digit and north of 40 percent response.

What Happens When You Ignore Signal Scoring

The alternative is spray and pray. Send the same template to 50 people, get two replies, and conclude LinkedIn doesn't work. It works. They're doing it wrong. This is the same quiet leak I describe in the cost of founder invisibility, only pointed inward at your outreach instead of your presence.

Every cold DM to an unscored connection does damage. You burn the first impression. You position yourself as someone who pitches strangers. And the algorithm notices. Low response rates signal low relevance, which depresses your reach. The cost of bad outreach includes the quiet penalty on your entire LinkedIn presence, not just the unanswered DM.

Scoring before outreach is the work itself. Founders who close from LinkedIn don't send more DMs. They send fewer, better-timed DMs the moment the signal says a connection is ready. Precision beats volume every time.

Building This Into Your Operating Cadence

The Compound Signal Score is a weekly ritual, not a one-time exercise. Every Monday, 30 minutes. Update scores for your pipeline universe. New engagement raises scores; decay lowers them. The list reshuffles. DMs go out. Engagement begins. The machine runs.

Founders who run this for 90 days stop sending cold DMs. Every message lands in a warm inbox because the score said it would. Response rates climb, meetings book, and deals close. The whole system runs on 30 minutes a week.

The VCO equation applies directly. Visibility (your content and engagement presence) multiplied by Time (consistent weekly scoring) multiplied by Relevance (targeting only high-fit connections) equals Opportunity Density. The Compound Signal Score is the layer that makes the equation measurable. I use the same logic when I build the founder pipeline dashboard: the score is the input and the dashboard is the readout.

Frequently Asked Questions

What is a compound signal score?

A compound signal score is a 0-to-10 measure of LinkedIn relationship warmth. It blends three weighted inputs: engagement depth (50 percent), signal recency (30 percent), and ICP relevance (20 percent).

How is it different from LinkedIn lead scoring?

Lead scoring grades a contact on fit and activity. The Compound Signal Score grades a relationship over time and across a buying group. It answers a different question. Is the relationship warm enough to message today, regardless of how good the fit looks on paper?

What score is warm enough to send a DM?

A 7 or higher. A 5 or 6 means you need one more engagement touchpoint before outreach. Below 5, don't DM. Build score first with the 7-day protocol.

How often should I re-score my connections?

Weekly, in one 30-minute session. Signals decay, and recency is 30 percent of the score, so a list that isn't refreshed drifts out of date within a month.

How do account-level compound signals work?

When multiple people from one company engage with your content in a short window, treat the account as warm. That holds even if no single person scores a 7. That pattern is a buying group researching in public.

Do I need to score all 2,000 of my connections?

No. Score the 100 to 200 high-relevance connections that form your pipeline universe. The rest stay on passive observation until a signal moves them into view.