Spotting LinkedIn Buying Signals and Reaching Out Warm
How to treat comments and likes on LinkedIn as buying signals, qualify engagers, and reach out warm instead of cold — including real reply-rate numbers.
A cold list knows who someone is. A signal knows what’s on their mind right now
Classic LinkedIn outbound works through lists — bought, scraped, or exported from Sales Navigator. Every person gets the same sequence, whether or not they currently have a problem. That works because volume works. But it overlooks the signal LinkedIn hands out for free every day: people publicly talking about exactly your topic area.
Someone commenting under a competitor’s post, someone asking a question about your product category, someone liking a post about their pain point — has, in that moment, shown more buying signal than any entry in a cold list. The difference between a comment and a like isn’t cosmetic: a comment is a formulated opinion, a like is pure attention. Weighting both the same dilutes your own signal.
Not every engager is a lead
The most common mistake in signal-based sourcing: treating the signal as sufficient. An engager initially only gives you a name, headline, and profile URL — raw data, not a lead. In between sits a qualification step:
- Resolve the company. The company and its domain are derived from headline and profile.
- Check ICP fit. Industry, size, and region against a fixed threshold, say 60 out of 100 points. Anyone below that isn’t contacted — no matter how strong the signal was.
- Determine the persona. A decision-maker and an intern respond differently to the same outreach.
Without this filter, students, freelancers, and companies outside your industry end up in the same sequence as your actual target customer — just because they happened to comment under the right post.
What sets a warm message apart from a cold one
The core mistake when starting out with signal-based outbound: disclosing the like or comment action itself. “I noticed you liked X’s post” comes across as surveillance, not relevance. The reference belongs on the topic, not the action:
- A short, substantive tie-in to the topic of the post.
- A genuine, easy-to-answer question — no pitch, no link in the first message.
- Only after a reply does the next step follow — never a direct meeting request in the first sentence.
Rate limits are also part of the methodology, not a footnote: LinkedIn tolerates roughly 20 to 25 connection requests per day and account. A connected account is a real account with a real ban risk — volume is deliberately not the goal here.
What practice actually shows
The honest number beats the hoped-for one. Our cold LinkedIn baseline sits at 26.2 percent reply rate across 6,605 evaluated messages. A customer running the methodology in production across several playbooks (Proseller AG) reaches 13 percent reply rate in the signal playbook that has run the longest so far, after a two-variant test — 3 replies out of 23 warm-contacted leads. One variant came in at 20 percent (2 of 10), the other at 7.7 percent (1 of 13).
That’s a small sample and not a solid claim that “warm beats cold” — on the contrary, the number sits below the cold baseline. The real value of signal-based sourcing isn’t a guaranteed higher reply rate — it’s access to contacts a cold list wouldn’t contain at all: people with a publicly visible, current interest in exactly your topic. Anyone setting up this method should treat it as an additional, highly relevant channel alongside existing outbound — not as its replacement and not as an automatic reply-rate boost.
When it’s worth it
Signal-based sourcing fits topics with visible, ongoing discussion on LinkedIn — enough post volume around your keywords to consistently find fresh engagers at all. For niche topics with little public discussion, the signal volume stays too small to build a reliable channel out of it; there, classic list-based outbound remains the backbone.
Next step
The full five-step methodology — defining keywords, scanning the feed, enriching, reaching out warm, running it as a loop, including two copy-pasteable prompts — is in the Academy playbook LinkedIn Keyword Monitoring. For signal-based outbound across channels beyond LinkedIn (website visitors, funding events), see Signal-Based Outbound.
If you’d rather not set up this loop yourself but have it run for you: check out GTM Goat or try it free for 4 weeks.
Common questions
What is a LinkedIn buying signal?
A publicly visible sign that someone is currently engaging with a topic that matches your offer — a comment under a post about that topic, a like on a competitor's post, a question in a group. Comments are the stronger signal because they show a formulated opinion; a like only shows attention.
Is everyone who reacts to a relevant post automatically a good lead?
No. An engager initially only gives you a name, a headline, and a profile URL. Only after company resolution, an ICP-fit check (industry, size, region), and a persona check against a defined threshold does that become a qualified lead. Without this filter, companies outside your industry or too small also end up in your outreach.
Does warm outreach beat the reply rate of cold outbound?
Not automatically. Our cold LinkedIn baseline sits at 26.2 percent (6,605 evaluated messages). At a customer running the methodology in production (Proseller), the reply rate on the signal playbook that has run the longest so far was 13 percent on 23 warm-contacted leads — a small sample, but below the cold rate. The value of warm outreach lies in reaching contacts with real, current interest, not in a guaranteed higher reply rate.
What makes a warm message different from a cold one?
It references the public topic the person spoke up on — never the like or comment action itself ('I saw that you liked...' comes across as surveillance) and never generic company data. A short, substantive reference plus a genuine, easy-to-answer question beats any pitch.
How many connection requests per day are safe on LinkedIn?
Roughly 20 to 25 per day and account is considered tolerable. A connected account is a real account — exceeding this limit risks a ban. Signal-based sourcing is inherently limited by the volume of fresh signals anyway; volume is deliberately not the goal here.