LinkedIn Connection Request: With a Note or Without?
A personalized note is accepted 45% to 60% of the time and a blank request 30% to 40%, so the note is worth roughly 15 to 20 points when it says something specific. A generic note is accepted 20% to 28%, below the blank request, which makes the templated note the only version that costs you acceptance. The measured half of this page covers what the connection is worth afterwards: once connected, our first messages were replied to 43.5% of the time, against 1.1% when the chat was open but no connection existed.
Last updated: September 2026 · Next update: December 2026 · Methodology
Compiled: acceptance rate by what you put in the request
| What you send | Avg acceptance rate |
|---|---|
| After they engaged with your post | 65–80% |
| Mutual connection mentioned | 60–75% |
| Reference recent post or comment | 55–70% |
| Personalized note, specific reason | 45–60% |
| Blank request, no note | 30–40% |
| Generic note, I would love to connect | 20–28% |
| Spammy template, Hi, I noticed you | 8–15% |
Compiled ranges, and the same seven rows published on our connection request acceptance benchmarks page. They are repeated here rather than restated differently, so the two pages give one answer.
What the note is worth
Personalized note against blank
+15 to +20 pts
Compiled. 45–60% against 30–40%
Generic note against blank
−10 to −12 pts
Compiled. 20–28% against 30–40%
Reply rate once connected
43.5%
Computed. 6,094 conversations, 13 accounts
Reply rate with no connection
1.1%
Computed. 186 conversations, 6 accounts
Computed: what the accepted connection is actually worth
| State of the relationship | Reply rate |
|---|---|
| Connected | 43.5% |
| Connection state not recorded | 34.0% |
| Chat open, not connected | 1.1% |
Outbound-initiated conversations, grouped by the connection state our system recorded for the thread. The bottom row rests on 6 sending accounts, below our floor of 8, and is published with that floor stated because the size of the effect matters: two replies from 186 messages. The middle row is conversations synced before connection state was tracked, kept visible rather than folded into either side.
Compiled: what a note has to contain to beat a blank request
| Element | Worth including |
|---|---|
| A named post of theirs | Yes, first line |
| One shared detail, city, event, employer | Yes |
| A mutual connection, named | Yes, if genuine |
| Your job title and company | No |
| A compliment with no specifics | No |
| A question or a call to action | No |
| A link | No |
Compiled guidance. Keep the note to one or two short sentences. Our dataset records conversations after they start, not the invitation that preceded them, so nothing here is a measured figure.
Why the blank request beats the templated one
A blank request carries no claim
The reader falls back on your profile and your headline. If both are relevant, that reads as a busy peer. There is nothing in the request to argue with.
A generic note makes a claim it cannot support
I would love to connect says the sender wants something and will not say what. That is why the compiled band, 20% to 28%, sits below the blank request at 30% to 40%.
Specificity is the whole mechanism
The three highest rows all name something real: a post, a mutual connection, a piece of engagement that already happened. What lifts acceptance is evidence that you did not send this to a thousand people.
Send blank rather than send generic
If there is no time to write something specific, send nothing in the note. That is the one place on this page where doing less performs better.
The acceptance is only the entry fee
Once the connection exists, our first messages were answered 43.5% of the time across 6,094 conversations. Messaging into a chat with no connection returned 1.1% across 186. The note buys the 43.5% door.
Frequently asked questions
Does a blank LinkedIn connection request really beat a templated note?
Yes. A blank request is accepted 30% to 40% of the time and a generic note such as I would love to connect is accepted 20% to 28%. A blank request reads as a busy, role-relevant person, while a templated note reads as a sales pattern. A specific personalized note beats both at 45% to 60%.
How long should a LinkedIn connection request note be?
One or two short sentences. Name one specific thing about them, give no ask, and stop. The pattern that fails is a note long enough to look like a pitch and general enough to have been sent to anyone.
What should a LinkedIn connection request note actually say?
One specific, checkable reason you are reaching out: a post of theirs you read, a mutual connection by name, an event you both attended. Leave out your job title, your company, any compliment without specifics, any question and any link. Those are the elements that turn a personal note generic.
Does the note matter if my profile is weak?
Less than fixing the profile does. A blank request is judged entirely on your headline and photo, and even a personalized note sends the reader to your profile before they decide. Fix the headline first, then write the note.
What happens after someone accepts a connection request?
In our data, first messages into an existing connection were answered 43.5% of the time across 6,094 conversations from 13 sending accounts, and 17.3% of those conversations reached three or more replies. The connection is what makes the message worth sending.
Is a connection worth more than messaging a non-connection?
By a wide margin in our data. Conversations where the chat was open but no connection existed returned a 1.1% reply rate across 186 messages from 6 sending accounts, against 43.5% once connected. That bottom row sits below our usual eight-account floor and is published with the floor stated, because the size of the gap is the finding.
Methodology
Acceptance figures are compiled and are the same seven approach bands published on our connection request acceptance benchmarks page. Our system records conversations rather than invitations, so we cannot measure acceptance and no figure here claims to.
Computed figures were run on 2026-09-09 across the 10,922 LinkedIn conversations in which our side sent the first message, from 17 sending accounts, drawn from 100,861 messages in 14,577 conversations.
Reply rate. Numerator: conversations in which the other person sent at least one message after our opener. Denominator: outbound-initiated conversations in that connection state. Reached 3+ replies uses the same denominator with a numerator of three or more messages back.
Connection state is what our sync recorded for the thread: connected, chat open without a connection, or not recorded. The not-recorded group covers conversations synced before the field existed and is published as its own row rather than being split between the other two.
The chat-open-without-connection row covers 186 conversations from 6 sending accounts, below our floor of 8. It is published with the floor stated on the page because the effect is large and the alternative is to hide it. Two of those 186 messages received a reply.
No conversation, contact, company or message text is exposed.
Compiled acceptance bands repeated from our own connection request acceptance benchmarks page and published B2B outbound benchmarks. Computed figures from Creator's anonymized LinkedIn DM dataset.
Understand the terms
Connection Request
Invite to add someone to your first-degree network.
Connection Note
The short message attached to a connection request.
Connection Acceptance Rate
Share of connection requests that get accepted.
Template Fatigue
The drop in replies once a message pattern is recognised.
Cold Outreach
First-touch outreach with no prior interaction.
Profile Conversion
How often a profile visit turns into an accept or a follow.
LinkedIn Content Playbook
The best note references something you published.
More LinkedIn benchmarks
See how your content compares
Creator tracks your performance against these benchmarks automatically and tells you exactly what to fix.
Apply to work with us