You accepted the connection. They accepted yours. Now there is a blank text box and a person who owes you nothing.
Most people freeze here, then paste something. The LinkedIn first message is the hardest one to write because it carries all the weight and has none of the context. There is no shared meeting, no referral, no prior email. Just a name, a job title, and whatever you managed to notice before you clicked send.
This article is about what goes in that box. Not a template to copy — the structure underneath, so you can write your own in under two minutes.
Why the first message is judged in four seconds
The recipient reads the first line in the notification preview. That is where the decision happens. Not in your third paragraph where the good part is.
They are asking one question, unconsciously: is this addressed to me, or to a list? Everything else — your offer, your credentials, your case studies — is irrelevant until that question resolves in your favour. If the answer is "a list", the message is archived without a reply and sometimes reported. Enough reports and you have a different problem entirely, which is covered in what gets a LinkedIn account restricted.
So the job of the opening line is narrow: prove a human wrote this, to this specific person, on purpose. That is it. You are not selling. You are passing a filter.
Five openings that kill a LinkedIn first message
These come up constantly. Each one signals "list" within four words.
"Thanks for connecting!" It says nothing, and everyone who pitches says it first. The reader has learned that a pitch follows. You have spent your opening line training them to skip.
"I see you're the Head of Operations at Acme." You have told them a fact they already know, pulled from a field anyone can read. It proves nothing about attention.
"I help [role] achieve [outcome] without [pain]." This is a positioning statement, not a message. It is about you. Opening with it means your first sentence contains zero information about them.
"Quick question —" followed by a paragraph that is not a question. People notice. It burns the one trick you had.
Any opening that works for 500 people. Test it: swap in a different name and company. If the message still makes sense, it was never personal, and the reader can tell from the texture of the sentence alone.
The common thread: all five are about the sender, or about nothing. The reader's attention is a loan and you have to pay interest in the first line.
The structure that actually works
Four parts. Roughly sixty to ninety words total. Longer than that and it looks like work to read.
One: the specific observation. Something you could only know by looking at them. A post they wrote. A job opening on their company page. A change in how they describe the team. A conference they spoke at. Not their job title. One sentence, no flattery.
"Saw you're hiring two more field technicians in the Bristol office."
Two: the inference. This is the part almost everyone skips, and it is the part that earns the reply. You say what that observation probably means for them. You are showing you understand their world, and you are giving them something to correct if you are wrong — which is itself a reason to respond.
"Usually that means scheduling starts breaking before the headcount actually lands."
Three: the relevance, in one line. Why you in particular are the person noticing this. Not your full offer. One clause.
"I spend most of my week on dispatch workflows for services teams that size."
Four: a question that is cheap to answer. Not "are you free Tuesday for 15 minutes?" You have earned a sentence of their time, not a slot in their calendar. Ask something they can answer in six words while walking.
"Is that what's happening, or have you got it handled?"
Read end to end, it takes ten seconds and reads like a colleague. No signature block. No link. No attachment. Nothing to click.
The part that makes it hard: finding the observation
This is where the method collapses in practice. People agree with the structure, then discover that writing the observation takes six minutes per person and they have forty people to contact.
Two things fix this.
Narrow the list until observations get easy. If your targets span five industries and four company sizes, every message requires original research. If they are all operations leads at 30-to-80-person field service companies, you start recognising the same three situations and the observation takes forty seconds. The narrowness is not a marketing exercise — it is what makes volume possible without becoming generic.
Separate research from writing. Do not look up a person and then write to them. Spend one block collecting observations into a list — name, one line of what you noticed — then a second block turning them into messages. Switching between research mode and writing mode is what burns the time. This is the same logic behind a 25-minute prospecting routine: batch the modes, not the people.
BIZLINKED is built around that split — you keep the research and the drafting in one place so the observation you found on Tuesday is still attached to the right person on Thursday. The writing stays yours. There is no version of this where a tool invents the observation for you, and messages generated at scale by software sit squarely in the territory covered by LinkedIn's rules on automation. Sending by hand is slower and it is the version that does not cost you the account.
What to do when they do not reply
Most of them will not. That is normal and it is not evidence the message was bad.
A good LinkedIn first message gets read, gets you remembered faintly, and gets no answer because the person was in a meeting when the notification arrived. Silence after one message tells you almost nothing. Silence after four tells you something.
So write the first message assuming it is the start of a sequence, not a one-shot. Two consequences:
- Do not front-load everything. If message one contains your offer, your credentials and your calendar link, there is nothing left to say in message two except "just bumping this."
- Keep the thread specific enough to resume. "Following up on the dispatch thing" works. "Following up on my previous message" does not, because they have no memory of it.
The follow-ups are where the meetings come from, and that is a separate craft — the logic of why the third message books more meetings is worth reading before you write your second one.
What to do Monday morning
Take the last ten people who accepted your connection request and that you never messaged. They are still warm enough.
For each one, spend ninety seconds and write one line only: what you noticed. Not the message — just the observation. If you cannot find one in ninety seconds, cross the person off. That decision is information: either the person is not really in your segment, or your segment is too wide to prospect efficiently.
Then take whatever survives — five, six people — and write the four-part message for each. Send them by hand, spread across the day.
Count replies, not meetings. At this stage a reply saying "no, we've got that covered" is a win: the message worked, the timing did not. Rewrite the observation line, not the whole thing. Nine times out of ten that is the part that failed.
