warmline
Free resource

The LinkedIn Comment System

A Claude prompt that writes thoughtful LinkedIn comments in your voice — comments that sound like you wrote them, and that get replies. Free. No email required.

Why comments work for B2B pipeline

A good comment on a buyer's post shows up in the feed of everyone who follows them — not just the person you commented on. It's the only LinkedIn action that puts you in front of the right people without paying for ads or cold messaging strangers.

When the comment is thoughtful and adds something, the person you commented on often replies. One reply is the start of a real conversation.

Why most people's comments don't work

"Great post!" and "Totally agree" get ignored. So does the comment that restates what the person already said. What gets a reply is a comment that adds a specific point, asks a genuine question, or offers a different angle — and sounds like a real person wrote it.

That's what this prompt is built to do. It uses your own voice examples so the output doesn't read as AI-generated.

How it works

01

Paste the post

Copy the LinkedIn post you want to comment on into the prompt.

02

Add two voice examples

Paste two comments or LinkedIn posts you've written that sound like you. The more different the topics, the better.

03

Fill in your context

Who you are, what you do, and the one thing you want this person to remember about you after reading the comment.

04

Run the prompt

Claude writes three options — different angles, same voice. Pick the one that fits the post best and edit it if you want.

05

Post it

Copy it to LinkedIn. The whole process takes under two minutes once your voice examples are saved.

Example outputs

Real comments written using this prompt on cybersecurity and B2B sales posts.

Post context

VP Sales at a Series B cybersecurity company posts about pipeline quality vs quantity

Comment written

The pipeline quality problem usually starts two stages earlier — the ICP definition is too broad, so every company with a security team makes it in. Once you tighten the filter, the rest of the stats improve on their own. How specific does yours go — persona-level or company profile?

Result

4 replies, 2 connection requests

Post context

Founder of an MSSP posts about winning a first enterprise deal

Comment written

The first enterprise win is the proof point that changes everything — not just for the clients who see it, but for how you price and position the next one. What changed in how you ran the deal vs your mid-market approach?

Result

Reply, 20-min call booked

Post context

Head of Compliance posts about the gap between NIS2 requirements and actual readiness

Comment written

The readiness gap is almost always a resourcing and prioritisation problem, not a knowledge problem — most compliance teams know exactly what needs to happen. The vendor side hasn't caught up with that yet. Are you seeing that in your pipeline too?

Result

3-comment thread, LinkedIn DM

The prompt

Copy this into Claude (claude.ai). Fill in the three sections marked with [brackets]. Run it each time you want to write a comment.

LinkedIn Comment System — Claude prompt
You are writing a LinkedIn comment in my voice. Your job is to write three different comments on the post below. Each comment should sound like I wrote it — not AI-generated, not generic.

MY VOICE EXAMPLES (two comments or LinkedIn posts I've written):
[Paste two examples of your own writing here — the more different the topics, the better the calibration]

WHO I AM AND WHAT I DO:
[One sentence: your name, what you do, and who you do it for. E.g. "I run an outbound agency for B2B cybersecurity companies."]

THE ONE THING I WANT THEM TO TAKE AWAY:
[What should they know or think about you after reading the comment? Keep it to one sentence.]

THE POST TO COMMENT ON:
[Paste the full LinkedIn post text here]

---

Write three comment options. Each one should:
- Add something specific that isn't already in the post (a counter-point, a related observation, a question that moves the conversation forward)
- Sound like the voice in my examples — same register, same sentence rhythm
- End with a genuine question when it fits naturally (not forced)
- Be between 40 and 120 words

Do not start any comment with "Great post", "This is so true", or any variation of agreement as the opening line.

Label them Option 1, Option 2, Option 3. No preamble or explanation — just the three options.

Get the full guide as a PDF

The complete LinkedIn Comment System — strategy, voice calibration, the prompt, 10 worked examples, daily workflow, and common mistakes. Free. No email required.

Download PDF ↓

Want this running automatically?

This prompt works well for one comment at a time. If you have 20–50 buyers you want to show up for consistently, doing it manually every day takes about two hours. Warmline does the same thing — finds your buyers' posts, drafts the comments in your voice, waits for you to approve — and tracks every reply, DM, and call back to the comment that started it.

Founding offerFirst 20 accounts lock 33% off, for life.