How to find leads in LinkedIn post comments

Open the post. Click the comment count underneath it. Keep clicking the load-more control until it stops appearing, and expand the replies nested under each comment. Every name is right there, and no tool is required to see them.

That is the entire mechanical answer, and most pages that rank for this question spend two thousand words on it before they stop. The reason they stop there is that the next part is genuinely hard and cannot be automated away: you now have a hundred names, and most of them are not buyers.

Why a comment thread fills up with the wrong people

Worth understanding as a mechanism rather than a complaint: the mechanism tells you what to discard.

A post gets shown to people whose recent activity resembles the activity of the people already reacting to it. So a post about a problem circulates first among people who talk about that problem, and the group that talks most about any business problem is the group that sells solutions to it. Vendors post about it, comment on it, and are shown more of it. They arrive ahead of the buyers, in larger numbers, and they comment more readily because commenting is their working day.

The second population is reciprocity traffic. Comments buy distribution, so a class of people comment on each other's posts as a standing arrangement. Their comments have a shape you learn to spot in a week: agreement with the general thrust, no reference to anything specific in the post, and often a closing pivot to their own view of the subject.

The third population is peers. If the author is a consultant writing about a problem their clients have, other consultants will show up to agree and add a refinement. They are not buyers. They are the author's competitors being collegial in public.

None of that is dysfunction. It is what a public thread is for. It just means the list you pulled is mostly not what you wanted, and the sorting is your job.

The one rule that does the work

A comment is worth something to you when it contains a situation. There are three forms it takes.

They asked a question. Somebody with the problem asks how the thing is handled when the awkward case comes up. The question itself is the evidence, because you do not ask about the awkward case unless you have met it.

They disagreed. Disagreement is the strongest of the three and the most overlooked. To argue with a post about a problem you have to have your own account of the problem, and people generally get that from living with it.

They described their own setup. Anything in the first person about how it currently works for them, including the version where they say it works fine.

Everything else is applause. Agreement without content, a compliment to the author, a single emoji, a tag of a colleague: none of these tell you anything, because none of them cost anything to write.

That is a rule about evidence, not about enthusiasm. One person who described their own month-end in two sentences is worth more than a hundred people who tapped a reaction, and the gap is not a matter of degree. The hundred left no evidence at all.

Worked through one thread

The thread below is constructed as an example, not a result from any account. The proportions are the shape to expect, not measured figures.

A fractional CFO writes about month-end close slipping because approvals sit in inboxes. The post collects 240 reactions and 34 comments. You sell approval workflow software to finance teams at companies of 50 to 500 people.

Six survive the rule out of thirty-four. Now apply the second filter, which is whether you can sell to them at all: of the six, four work in finance at companies inside your size band, one is at a 4,000-person group where the decision is three layers away, and one is a student.

Four messages from thirty-four comments. Against the 240 reactions, which produced 240 names and not one sentence, the arithmetic is four out of two hundred and seventy-four people who touched that post. Under two percent.

The first reaction to that number is that the effort is not worth it. Reading the thread properly takes about twelve minutes. Four people you can write a specific first line to, in twelve minutes, beats the hourly rate of most outbound work, and the alternative is sending 240 identical invites for nothing except a slightly worse reputation with 240 people.

Why reactions cannot be rescued

People keep trying to make reactions work, usually by counting them. Someone who reacted to five posts about the problem must care about the problem.

The trouble is that a reaction is a single bit and the reason behind it is unrecoverable. It can mean agreement, or support for the author regardless of subject, or a wish to be seen in the thread, or two interesting lines and a scroll. Those are different people, and the tap looks identical for all of them.

What a reaction does tell you is that the post reached that person's feed and they did not scroll past. That is a fact about how the platform distributed the post. It is not a fact about their situation, and the ranking method in how to prioritize sales leads puts it exactly where it belongs, at the bottom, with the other category matches.

There is one honest use. If you already have a reason to think a specific person has the problem, seeing them react to a post about it is a weak confirmation. It is a check on a name you already had. It is never a way to find one.

Competitors' threads, and why they are different

A competitor's post is the densest source available, because the people who show up already accept that the category exists. You skip the whole argument about whether the problem is worth solving.

The population changes shape, though. Vendors thin out, because rivals mostly do not comment on each other in public. In their place you get the competitor's own staff, their partners, their happiest customers doing them a favour, and their evaluators.

The sorting question becomes what the person is to the product rather than what they said. Someone asking how the pricing works past a certain volume is evaluating. Someone describing a limit they ran into is using it and unhappy. Someone congratulating the company on a milestone is a peer or an employee. The first two are worth your time and the third never is, no matter how senior the name looks.

Be careful with the unhappy user. A person publicly complaining about a competitor is not automatically ready to move, and a message that arrives an hour later reading as though you were waiting for them to complain does not land well. Read what they actually said about the limit, and write about the limit.

The part nobody mentions: sourcing and writing are the same step

The reason this is worth doing by hand is not that the names are better. It is that by the time you have decided someone is worth contacting, you have already written the first line of the message.

The comment that qualified them was a sentence about their situation. That sentence is the reason you are writing. There is no second research pass, no hunt for something personal to open with, no template with a variable in it.

Three things not to do with it. Do not put their words in quotation marks, because that reads as a mail merge whether or not it is one. Do not open by saying where you found them, because that makes your looking the subject of the message instead of their problem. And phrase it as a question rather than an assertion, because you are working from one sentence in a comment thread and you may well have misread the situation behind it. How to say why you are reaching out goes through the line itself in more detail.

Where this does not work

If your buyers are not on LinkedIn, none of this applies and no method fixes it. Deskless trades, most local service businesses, large parts of the public sector: the threads you would need to read do not exist, and reading the wrong threads harder will not conjure them.

If you need volume, this is the wrong tool. Twelve minutes a thread and four names is fine for someone who wants ten good conversations a month. It does not become fifty conversations by working faster, and a bought list will not close the gap either, for the reasons in should you buy a lead list: the names arrive without the sentence that made them worth contacting.

If you sell something cheap into a very large market, where a contact costs you nothing and the decision takes a minute, sorting carefully costs more than it saves. Volume genuinely wins in that case.

Doing it at any scale

The manual version has a hard ceiling. You can read maybe five threads a day around actual work, and you have to find the threads first, which is the slower half of the job.

That ceiling is why Trendle exists. It reads threads at volume, drops the reaction-only names, and keeps the people who left a sentence, so what arrives is names with the sentence attached rather than an export of everyone who touched a post. What it does not do is decide whether the situation someone described is one you can solve. That judgement depends on what you sell and on things you know that no software does, and the four-out-of-thirty-four arithmetic does not improve when you automate it. It just stops costing you the twelve minutes.

The number worth tracking, whichever way you do it, is what share of the names on your list have a sentence next to them that the person wrote. If that share falls while the list grows, you are collecting names instead of reasons, and replies follow the share rather than the size.