Note or no note: a segment decision, not a copywriting decision
By Jānis Plūme, Founder, Outbound Pros and LinkedPros · 2026-08-06
Quick answer
Whether a LinkedIn connection request carries a note is decided by the segment it is going to, not by how good the note is. Cold, high volume, well matched segments run blank by default on our client campaigns, because a blank request from a credible peer asks for a smaller decision. Segments where a specific, checkable reason to connect actually exists earn a note, and that is where writing it well pays. How to write one belongs to the parent site. What is on this page is how to pick the mode before you send, and what it costs you when the mode and the segment do not match.
Should a LinkedIn connection request include a note?
It depends on the segment, and that is a real answer rather than a hedge. A note is worth its cost when the segment supplies a specific, checkable reason the two of you should be connected. It is a liability when it does not, because a note with nothing specific in it converts an ordinary networking request into a sales decision before the recipient has received anything.
Here is the mechanism, which is the part that makes the segment rule follow. When a request arrives, the recipient sees a name, a headline, a photo, a shared connection count and, if you sent one, a note. They decide in seconds. A blank request from someone who looks like a relevant peer reads as ordinary networking, so it asks for a small yes. A note that announces intent asks for a much larger one at the moment of least trust. That trade only reverses when the note carries information the recipient did not already have from the profile preview.
The default we run on client campaigns is blank, with the persuasion moved into the profile and the first message after acceptance. That is our operating default and not a universal law. We report it separately from the 59% acceptance figure on purpose, because we have not confirmed whether the control cell behind that figure ran blank notes, and putting the two facts in one sentence would imply a link we cannot evidence.
Which segments earn a note, and which do not?
Four segment shapes cover almost every list we build. The mode follows from the shape, and the decision takes about a minute per segment once the list is defined.
| Segment shape | Mode we run | Why | What it costs you |
|---|---|---|---|
| Cold, matched by industry and seniority, sender credible to that audience, shared connections visible | Blank | The profile preview already answers why is this person contacting me | Slightly less qualified accepted connections, because the yes was cheap |
| A real named mutual contact, a shared event, a piece of their published work you genuinely read | Note, written properly | The specific fact is the entire reason acceptance moves | Time. This mode does not scale and should not be asked to |
| Senior buyers, cold, no overlap you can point to | Blank, and expect a lower rate | Seniority responds to who is asking, and a note cannot change who is asking | A rate your plan has to accept in advance |
| Broad list assembled by job title across several geographies | Neither. Fix the list | Both modes underperform on a list that was not built to fit the sender | A week you will otherwise spend rewriting |
The fourth row is the one nobody wants. If the segment is wrong, the note question is not live yet, and answering it is how three weeks disappear.
Never send a note that opens with flattery about their post, invents a shared connection, mentions their company's impressive growth, or asks for time. Fake personalization is worse than none, because it signals a system while pretending to be a person. The version of that argument with the actual sentence patterns in it is on the Outbound Pros blog, which owns connection request copywriting.
Why does the segment decide more than the note does?
Because the recipient is judging who you appear to be relative to them, and the note cannot change that in the two seconds available.
Run identical request text against two lists. One is people in your sender's industry, at comparable seniority, in a market where your company means something. The other is a broader list assembled by job title alone. The gap between those two acceptance rates is wider than the gap between a good note and a bad one on either list.
That is why the note question is downstream. When a rate falls, the check order runs targeting drift, list exhaustion, sender mismatch, send velocity, and only then copy. The full diagnostic with the check order and the cost of each check is on acceptance rate benchmarks, which is where that method lives. If the underlying problem is how the list was built, the group's lead generation work is the right place to look, not another copy round.
How do you pick the mode before you send instead of after?
Four questions, answered per segment, before a single request goes out.
- Can you name a fact about this segment that is true of every person in it and not true of your other segments? If yes, that fact is the note. If no, blank.
- Is the sender credible to this specific audience without explanation? A founder writing to founders needs no note. A sales title writing to founders cannot fix that gap with one.
- Will shared connections show on the request? They do more reassurance work than any sentence you can write, and they are visible before the note is read.
- What rate does the plan assume for this segment? If the plan needs a rate a senior cold list has never produced, the mode is not the problem and no note fixes it.
Write the answers down next to the segment definition. On inherited campaigns the most common finding is that nobody ever recorded which mode ran against which segment, so nothing that happened afterwards could be attributed to anything.
What breaks when the mode and the segment do not match?
Two failures, and both get misread as copy problems.
A note on a segment with nothing specific to say. The note has to fill space, so it fills it with the sender's own agenda, and the request reads as bulk. Acceptance drops, the team rewrites the note, and the rewrite has the same nothing to work with.
Blank on a segment where the reason to connect was the only asset. A real mutual contact goes unmentioned, so the request lands as generic against an audience that would have accepted a specific one. The cost here is invisible, which is what makes it worse: the rate looks acceptable and the best segment on the list quietly underperforms.
There is a third failure that is not about mode at all. When a rate slides gradually with nothing changed, the usual cause is that the good fit segment has been worked through and the remaining volume is drawn from worse fit prospects. Plan for exhaustion before it happens: know how large the good fit segment is at the start, and have the next segment or the next angle ready. This is where WideNET angle testing earns its place, because a framing already tested across the market is one you can deploy on the next segment instead of discovering under pressure. Sending faster than the account should also produces a drop that looks like copy and is pacing, and those mechanics are on connection request limits.
Frequently asked questions
Is a blank connection request rude?
No, it is the platform's default behaviour and the majority of invitations on LinkedIn arrive without a note. The perception that a note is more polite comes from email norms, where an unexplained message is genuinely strange. A connection request is a different object with a different convention.
Does the note character limit matter?
Less than people think, because almost every note that works is far shorter than the 300 character limit. That figure is settled and the parent site publishes it in its connection request guide, but we do not carry a verification date for it here, so treat it as a limit to check rather than a limit to cite. If you are running out of room, the note is doing work that belongs in the first message.
Does viewing someone profile before connecting increase acceptance?
We have not run a controlled test on viewing before connecting, so this answer is reasoning and practice, not measurement. The mechanism people cite is that a view creates a notification, the recipient looks back, and the request arrives to a name they have already seen. That is plausible and it is what we do by default. Every published percentage we have found on it comes from a company selling the tool that does the viewing, so we are not adding another unsourced number to that pile.
How do you get founders and CEOs to accept?
Send from a peer, be findable, use warmth you actually have, and plan for a lower rate. Seniority responds to who is asking and to whether the profile stands up to a closer look, which senior readers give it. The sending account is a strategic choice, not an administrative one, and on client programmes it is one of the first things we change.
Should I mention my company in the note?
Only if the company name is the reason they would accept, which is true for recognisable brands and false for everyone else. For most senders the company name converts a networking request into a sales request, which is the trade in the first section.
What do I write in the first message after they accept?
A different job with different rules, and the one place in this channel where a rewrite reliably changes the outcome. It is covered on our LinkedIn DM sequences page.
Do you use AI to write connection request notes?
We use AI for enrichment and research inputs, not for manufacturing personalization that does not exist. AI written notes tend to produce a sentence about the prospect that is technically derived from their profile and obviously generated, which lands worse than no note. When we run this for clients the client approves the messaging before anything sends.
When we run this at volume it is our managed LinkedIn work, and the same discipline sits behind everything published at the agency behind this site.
Last updated: 2026-08-06
Know your ceiling
before you hit it
The calculator returns a safe daily figure and a ramp schedule for your accounts. It runs in the page and sends nothing anywhere.
Free. No signup, no email capture.