On this page
- You are borrowing a comment section, not an audience
- The twenty-account rule selects for reach you have no way to convert
- The Ra-Aha Overlap Screen: five observable tests, scored out of ten
- The overlap arithmetic, with 6,000 right followers against 90,000 wrong ones
- Build the shadow list, because the real targets are the regulars
- What the habit costs, and why fewer comments in better rooms wins
- The thirty-day read: three numbers, each able to come back negative
- When the read comes back flat, check these five things in this order
- What commenting cannot do, stated plainly
Commenting works as distribution because you borrow a comment section, not an author. So select rooms by who is standing in them rather than by how many followers the host has. Score each candidate on five observable tests, name six buyer-side regulars per room before you commit, and read the result at thirty days on the job titles of your profile viewers. Comment volume is the input everyone optimises and the one that matters least.
You are borrowing a comment section, not an audience
The reach of your comment is capped by the reach of the thread it lives in, and the thread is a smaller and much more specific object than the author's follower count suggests. A host with ninety thousand followers who publishes something that collects eleven comments has produced a thread that went nowhere, and being the twelfth voice in it buys you nothing. A host with six thousand followers whose posts reliably gather forty replies has built a room that people actually stand in.
This reframe changes the selection problem completely. If you are borrowing an audience, size is the obvious criterion and every guide on this subject is right to tell you to find accounts two to ten times your own. If you are borrowing a room, the criterion is who is in the room, and size stops being informative almost immediately.
The rest of the standard advice survives this reframe intact. Commenting inside the first hour is sound, because early comments sit higher in the thread and the thread is still being distributed. Five to ten comments a day is a reasonable working volume. Neither of those rules tells you which twenty accounts to spend a year on, and that is the decision that determines whether the other rules produce clients or applause.
of members post more than once a week. The layer of the platform that speaks out loud is thin, which is the whole reason recurring presence in one comment section registers as a person rather than as a notification. You cannot achieve that in a feed and you can achieve it in a room of forty regulars.
Aggregate 2026 LinkedIn statistics reportsThe twenty-account rule selects for reach you have no way to convert
Follower count is chosen as the selection criterion because it is the one number sitting on the screen, not because it predicts anything you care about. It is a stock measure of an audience accumulated over years. What you need is a flow measure of how many of the right people read a given thread this week, and that number is observable too. It is just two clicks further away.
The observable substitute is thread density: the median comment count across a host's last ten posts, expressed per thousand followers. An account with 6,000 followers and a median of 40 comments has a thread density of 6.7. An account with 90,000 followers and a median of 25 comments has 0.28. Both numbers take about four minutes to gather and one of them tells you whether the audience is present or merely counted.
The second observable is composition, and it is gathered the same way you would audit your own audience. Open the host's last three threads, read the headlines of the thirty most recent commenters, and sort them. If you have already run the exercise in the audience census for your own posts, this is the identical method pointed at somebody else's room, which makes the two numbers directly comparable.
The Ra-Aha Overlap Screen: five observable tests, scored out of ten
Score every candidate account zero, one or two on each of the five tests below. Keep the accounts scoring seven and above and discard the rest without sentiment, including the ones you personally enjoy reading. Every input here is visible from the outside in under ten minutes per account, which means the whole screen for forty candidates is one long afternoon.
Anything scoring below seven is a hobby rather than a channel. The uncomfortable part of running this screen properly is that several of the largest accounts in your field will score two or three, and you will have to drop them while watching other people keep commenting there.
The overlap arithmetic, with 6,000 right followers against 90,000 wrong ones
Two accounts can hold almost the same number of your buyers while differing by a factor of fifteen in headline size. The calculation below shows how, and every figure in it is either observable in four minutes or an assumption you should replace with your own reading of the room.
| Measure | Account A, chosen by size | Account B, chosen by overlap |
|---|---|---|
| Followers | 90,000 | 6,000 |
| Median comments, last ten posts | 25 | 40 |
| Thread density per 1,000 followers | 0.28 | 6.7 |
| Estimated buyer share of the room (your reading) | 2% | 35% |
| Buyer-side people in the wider audience | 1,800 | 2,100 |
| Overlap Screen score | 3 of 10 | 9 of 10 |
Illustrative comparison. Followers and median comments are observable. The buyer-share estimates come from classifying thirty commenters, which is a reading you take yourself, not a published figure.
On raw buyer-side audience the two accounts are effectively tied. On whether that audience shows up to read the thread your comment is sitting in, they are not close. The small account wins on both dimensions at once, and it wins for a structural reason rather than a lucky one: rooms stay dense when they stay specific, and audiences get diffuse as they get large.
Nobody publishes the buyer share of another person's comment section, and any number you are handed for it is invented. The 2% and 35% above are illustrative. The method for producing your own is not: read thirty headlines, sort them, divide. It takes about six minutes per account and it is the only part of this screen that cannot be skipped.
- A comment's reach is bounded by the thread it sits in, so the host's follower count is a weak predictor and the host's median comment count is a much better one.
- Three comments a day in rooms with high buyer overlap can produce several times the buyer-side impressions of eight comments a day in rooms picked by size, at half the hours.
- You are choosing twenty comment sections rather than twenty authors, and the test of a good choice is whether you can name six buyer-side regulars in each room before you start.
- The standard advice is unmeasurable as written, so the thirty-day read has to be defined in advance around three numbers that can each come back negative.
- Thirty days reads composition and never reads revenue, because a comment creates recognition rather than urgency, and recognition converts on the buyer's timetable.
Build the shadow list, because the real targets are the regulars
For every account that passes the screen, the actual target is not the host. It is the ten to thirty people who keep turning up in that host's comments. Name them, write them down, and the entire exercise becomes countable, because you defined the list before you started and cannot retrofit success onto it afterwards.
Across the host's last five posts, note every person who commented more than once. Most active rooms produce between ten and thirty recurring names. These are the people who will see you repeatedly, because they are in the room as often as you plan to be.
Classify the regulars by headline and keep the ones who could sign. If a room yields fewer than six buyer-side regulars, you selected the author rather than the room, and the account should not have passed test one. Send it back.
Your comment has one job, which is to be useful to the six named people. The host is the venue. This changes what you write more than any tip about comment length: you stop performing agreement for the author and start making a distinction the regulars can use.
A reply from one of your six is the event that counts. A like from the host is pleasant and means nothing. Because the list is finite and named in advance, this is a real number that can come back as zero, which is exactly what makes it worth tracking.
If thirty days pass and not one of the six has ever viewed, followed, replied or connected, swap the account out. Do not extend the trial because you like the host. The rotation rule is what stops a commenting habit from calcifying into twenty rooms you visit out of loyalty.
There is a simple test for whether you have done this properly. Before you commit a year of mornings to an account, you should be able to name six people in its comment section from memory. If you cannot, you have chosen a person to admire rather than a room to work in.
What the habit costs, and why fewer comments in better rooms wins
Eight comments a day at six minutes each, which includes reading the post properly rather than reacting to the first line, is forty-eight minutes a day. Five days a week across forty-eight working weeks is 192 hours a year. Priced at a self-assessed rate of 150 dollars an hour, purely as an assumption you should swap for your own, that is 28,800 dollars of your time. The selection question is therefore a 28,800 dollar question, and most people answer it by glancing at follower counts for ten minutes.
| Protocol | Comments per day | Assumed readers per thread | Assumed buyer share of room | Buyer-side impressions per day | Minutes per day |
|---|---|---|---|---|---|
| Volume, rooms chosen by size | 8 | 300 | 2% | 48 | 48 |
| Selective, rooms chosen by overlap | 3 | 300 | 35% | 315 | 24 |
Two protocols compared on buyer-side impressions per day. Thread readership and buyer share are assumptions. The structure of the calculation is not.
A third of the volume produces more than six times the buyer-side impressions, at half the daily time cost. That result is not a claim about how well anyone comments. It falls straight out of multiplying volume by composition, and it holds for any pair of numbers where the overlap gap is wide enough. The binding constraint on this channel was never how many comments you leave. It is how many good rooms exist in your market, and the honest answer is usually somewhere between eight and twenty.
This also resolves the most common objection to commenting, which is that it does not scale. Correct. It does not scale, and the reason is that room quality is the scarce input. Doubling your comment count inside the same twenty rooms mostly increases how often the same regulars see you, which has a real and rapidly diminishing return.
The thirty-day read: three numbers, each able to come back negative
Define the read before you start, or the habit becomes unfalsifiable in exactly the way the published advice already is. Three numbers, all of which need a baseline captured on day zero, and each of which rules out a different explanation when it fails to move.
| Number | How to capture it | What a flat reading proves |
|---|---|---|
| Buyer share of your profile viewers | Classify the job titles of everyone your account shows you as having viewed you, in the thirty days before and the thirty days after | If total views rose and buyer share did not, you selected rooms by size. The Overlap Screen scores were wrong, not your comments |
| Named-target contact count | Count how many of your shadow-list regulars viewed, followed, connected, replied or wrote to you | Zero across twenty rooms means either your comments are not visible or the rooms are wrong. Check visibility first because it is cheaper to fix |
| Second-order replies | Count replies to your comments from anyone other than the host | Zero means your comments are agreeable rather than useful. Nobody replies to agreement, and agreement is what most commenting advice produces |
The thirty-day read. Capture the day-zero baseline before your first comment or the comparison is void.
Note what is deliberately absent from that table. There is no revenue line, because thirty days cannot produce one honestly. A comment does not create urgency in a buyer, it creates recognition, and recognition converts when the buyer's own timing arrives rather than when your quarter ends. The earliest sensible revenue read on this channel is ninety days, and anyone selling you a thirty-day pipeline result from commenting is selling a tool.
When the read comes back flat, check these five things in this order
A flat thirty-day read is information, not failure, and the order of investigation matters because the cheapest repairs sit at the top. Work down the list and stop at the first thing that is clearly broken rather than changing everything at once.
- Visibility. Open three of your own recent comments and note where they now sit in the thread. If you are consistently below the fold, the first-hour rule is not being met and nothing else you change will matter.
- Shape. Read your last twenty comments and count how many contain a claim somebody could disagree with. A comment that adds a distinction earns replies. A comment that agrees earns a like and leaves no trace.
- Room composition. Re-read the thirty most recent commenters in each room and re-score test one. Rooms drift, and an account that scored nine in January can be a peer convention by June.
- The destination. If named targets viewed your profile and nothing followed, the room worked and the landing did not. That is a separate repair, described in the profile as a conversion step.
- The next move. If you commented for thirty days and never wrote to anybody, recognition simply decayed. Commenting warms a named list. It does not contact them for you.
That last point is where most commenting programmes quietly end. The regulars know your name, they have seen you be useful eleven times, and then nothing happens because no one decided who was supposed to speak first. The follow-through for a small named list looks nothing like sequence outreach, and it is covered in writing directly to named executives and in what to say after they accept.
What commenting cannot do, stated plainly
Commenting cannot create demand. It places you in the recall set of people who already have the problem, so that when their situation changes your name is one of the two or three that surface. That is a real and valuable function and it is a smaller claim than the one usually made for this tactic.
- It cannot rescue a profile that fails inspection, because every successful comment sends people to a page you have not fixed.
- It does not compound with volume. Room quality is the scarce input, and there are only so many good rooms in any market.
- It is not free distribution. It costs between 96 and 192 hours a year on the arithmetic above, which is a real line item even though it never appears on a budget.
- It produces recognition rather than urgency, which means the conversion event is decided by the buyer's calendar and not by your consistency.
There is one failure mode specific to doing this well. If commenting is the only place your thinking appears, you become known in the room as a commenter, which is a real identity and a slightly diminished one. The correction is not to comment less. It is to make sure that a regular who follows you home finds a body of work rather than a body of replies, so the recognition you built in someone else's room has somewhere of your own to land.
monthly visits to the platform, measured in February 2026. Almost none of that traffic is reachable by you. The twenty rooms on your list are, which is the entire argument for treating selection as the strategic decision and comment volume as an implementation detail.
Semrush, 2026Questions people ask next
How many accounts should be on the list at once?
Does commenting inside the first hour really matter?
Should I comment on posts by my direct competitors?
How long before commenting produces an actual client?
Is it worth using a tool that drafts comments for me?
Send your profile, keep the aha.
One email with your LinkedIn URL. One specific, actionable insight back, free. If it is obvious, we did not look hard enough.
Email your profile →