If you want to automate LinkedIn messages without sounding like a robot, the fix is small. Before you send, find one real detail in the person's recent activity and refer to it in your first line.
A post, a comment or a job change is enough. It tells the reader a person looked at their profile, and that is what most people check for in the first five words.
This guide shows where to find that detail, how to write the rest, and what to check before anything goes out.
Why do automated LinkedIn messages sound like robots?#
Most of them start the same way: "Hi {first_name}", then a line about their job title, then a pitch. The reader has seen this a hundred times and can spot it at once.
The problem is not the automation. It is that the message could have been sent to anyone. If a message works for a thousand people, the reader knows it was written for none of them.
A mail-merge field swaps a name or a company into a fixed text. It changes the label, not the message. A person reading it can tell.
What counts as a real detail?#
A real detail is something only this person could have received. It comes from what they did lately, not from a field in a database.
Good places to look:
- A recent post they wrote, and what they said in it
- A comment they left on someone else's post
- A job change or new role
- A topic they keep returning to across several posts
- Something they shared that shows what they care about this quarter
Skip details that are only facts about the company. "Congrats on being in the software industry" is not a detail. A line about a specific point they made last week is.
How do you write the message around it?#
Put the detail in the first line. Then write the rest the way you would say it out loud: short, plain, one ask.
- Open with the detail. One sentence that shows you read their post or noticed their move.
- Say why you are writing. One sentence. Link it to the detail if it honestly connects.
- Make one ask. A question they can answer in a few words, not a request for a call on the calendar.
- Stop. No second pitch, no list of features, no "just circling back" in the first message.
Here is an example of the shape. The facts are made up for illustration:
"You wrote that most onboarding emails get skipped because they arrive too early. That matched what we see with small sales teams. Do you handle follow-ups yourself, or does someone else own them?"
It is three sentences. It refers to something specific, and it asks one thing.
What should you check before you send?#
Run every message past a short checklist. If it fails one item, rewrite it.
- Could this first line be sent to someone else? If yes, it is not specific enough.
- Does the detail come from something they actually did, not a guess?
- Would you say this sentence out loud to them at an event?
- Is there exactly one ask?
- Is it short enough to read on a phone without scrolling?
Also read it as the recipient. If it sounds like an advert, cut the second half. Our piece on writing connection notes that do not feel automated covers the shorter invite note in the same way.
What should you do this week?#
Try this by hand first, so you know what good looks like before you automate any of it.
- Pick ten people you want to reach.
- Open each one's recent activity and note one post, comment or job change.
- Write a first line from that note, and one ask. Keep each message under a few short sentences.
- Send them, and keep a note of which details drew a reply.
- Plan the follow-up. Our three touch LinkedIn sequence gives a simple structure for it.
The ten messages will take time. That time is the point: it shows you what the research step costs, and why it is the part worth handing off.
How does LinkedMinds do this for you?#
Doing this by hand for every contact is where the hours go. LinkedMinds estimates 2 h 35 m of manual outreach a day, from 20 personalised touches, and replaces that work.
Each message is written from the recipient's real profile and recent activity, in the sender's voice. There are no mail-merge templates. The detail in the first line comes from the person, not from a field.
Pacing and safety are handled too. Actions are spaced with randomised 2-10 minute human pacing, and the account has a 6-layer safety net. You can read the details on the security and safety page.
The aim is not volume. It is fewer, better messages that read like a person wrote them, sent in a way that protects your account.
Frequently asked questions#
Can you automate LinkedIn messages and still sound personal?#
Yes, if the automation reads each person's real profile and recent activity and writes from that. It does not work if it only swaps a name into a fixed text. The test is whether the first line could go to anyone else.
How much personalisation is enough?#
One real detail is usually enough. A single specific line about a post, comment or job change shows you looked. More than that can start to feel forced.
Should every message mention a recent post?#
Not always. Some people have no recent activity, so a job change or a shared topic can work instead. If there is nothing real to refer to, it is better to wait than to invent a detail.
If you want messages that read like a person wrote them, without spending your afternoons on research, you can start a free trial. It runs for 15 days and needs no card.
