Writing the First Message
What it does: turns an accepted connection into a conversation. This is the message that decides your reply rate.
Run it when: every campaign, after the connection request is settled.
Time: 30 minutes, and worth every one of them.
Before you start
One real proof point, with a number, from the brand profile
The offer sentence from SOP 02
A decision on whether you are using a per-lead AI opener, which is SOP 12
The steps
1. Remember what just happened.
They accepted. This is not a cold DM to a stranger, it is the first thing you say to someone who already said yes. Writing it as a cold pitch wastes the one advantage you have.
2. Use the three-line shape.
Personalized opener, one trust line, one soft question. That is the whole message.
Hi {{first_name}}, {{ai:first_line_template}}.
We booked 47 meetings in 30 days for a three-person sales team using two campaigns.
Would you be open to a quick chat about what we are seeing work right now?3. Keep it under 60 words.
Two to four short lines, one sentence per line, readable on a phone in five seconds. The field allows far more; using it is the mistake. Length is the most common reason a good message gets no reply.
4. Give one proof point, never a feature list.
One concrete result with a real number beats everything. If the brand does not have one, walk down the ladder honestly: a status or authority line, then a specific observation about the buyer's world that proves you understand it, then a soft benefit claim as a last resort. Never invent a result, a client name or a statistic. Inventing proof is the fastest route to being reported.
5. Ask a question, not for a sale.
The goal is a reply. "Would you be open to..." and "Curious how you are handling..." get answers. "Book 15 minutes here" gets ignored, and it needs a link, which brings us to the next step.
6. No links. Not one.
Links in cold LinkedIn messages read as spam and sink reply rates. The link goes out after they reply, which is SOP 16, Handling a Reply and Booking the Call. The only exception in the whole system is the delivery link in a lead-magnet DM, which the person explicitly asked for by commenting.
7. Make sure it survives losing the AI line.
If you use a per-lead AI opener and the generation fails, the tag is removed cleanly and the rest of the message sends as written. So the trust line and the CTA have to stand on their own, and nothing after the opener may refer back to it with a "that's why" or a "given that".
8. Vary the shape between campaigns.
If every campaign in the workspace opens the same way, the tenth prospect who has now seen three of your senders recognises the pattern. Same voice rules, different structure.
Do this / Not this
Do this | Not this |
|---|---|
Under 60 words | 200 words of context |
One result with a number | Three features and a value proposition |
A question that is easy to answer | "Here's my calendar link" |
One sentence per line | Dense paragraphs |
Honest proof, or none | An invented case study |
Copy that still works if the AI line drops | "Given that, we thought..." |
Plain human words | "Excited to connect and leverage synergies" |
Done when
Under 60 words, two to four lines, one sentence each
Exactly one trust line, and it is true
CTA is a low-pressure question
Zero links
Reads complete with the AI opener removed
Passes the humanize test read aloud
Every merge tag reads correctly when the underlying value is missing
