What is the best LinkedIn message template for cold outreach?

There isn't one, and any page claiming otherwise is selling you a template that thousands of other people are already sending. What consistently works is a structure: one line that proves you looked at them specifically, one line of relevance to their situation, and one low-friction question. Under 400 characters, no pitch, no calendar link in the first message.

Should I add a note to my LinkedIn connection request?

Only if you can make it specific to that person. The published data disagrees about notes, but not in the way most articles suggest: on acceptance rate the difference is small either way, with the largest study finding 26.42% with a note against 26.37% without. What the studies do agree on is that a generic note is worse than none, and a genuinely specific one is better than either. If you can't fill 300 characters with something true only of that person, send it blank and spend the personalization on the first message instead.

How long should a LinkedIn message be?

Aim for under 400 characters for a first message and under 300 for a connection note, which is LinkedIn's hard cap. On a phone, roughly four lines is where a message stops being read and starts being skimmed. Messages of one or two sentences reliably outperform messages of six.

How many follow-ups should a LinkedIn sequence have?

Two or three after the first message, spaced three to five days apart, then stop. Reply rates fall off sharply after the third touch, while complaint and block rates keep climbing. A sequence with seven follow-ups doesn't convert more people, it just annoys more of them and damages the sending account.

Do personalized LinkedIn messages actually get more replies?

Real personalization does. Merge tags do not. {{firstName}} and {{company}} are visible to recipients as automation, because everyone gets the same message with their own name in it. The personalization that moves reply rates references something only true of that person: a post they wrote, a role change, a specific detail from their company's site.

What is the best time to send LinkedIn messages?

Tuesday to Thursday, mid-morning in the recipient's timezone, is the commonly cited window. Treat it as a tiebreaker rather than a lever. It moves reply rate by a few points at most, while targeting and message quality move it by multiples. Sending outside business hours matters more as an automation signal than as a reply-rate factor.

Should I send a LinkedIn InMail or a connection request?

Connection request first, in almost every case. InMails cost credits, land in a visually distinct inbox that experienced buyers have learned to ignore, and carry no relationship context. InMail earns its place when the person is unreachable otherwise: a closed profile, or a role senior enough that their invite queue is permanently full.

How do I follow up without being annoying?

Add something each time instead of asking again. A useful follow-up brings a new angle, a relevant example, or a piece of information they can use whether or not they buy. 'Just bumping this up' asks the recipient to do the work of remembering why they should care, which is why it converts so badly.

LinkedIn Message Templates That Get Replies (2026), Outvelo

The short version

A note only helps if it's specific. Generic note < no note < specific note.
First message under 400 characters, no pitch, one easy question.
Merge tags are not personalization. Recipients read {{firstName}} as a robot.
Two or three follow-ups, then stop. Add value each time.
Rewrite every template below. The reasoning is the reusable part.

Every template on this page will stop working eventually. That's not false modesty, it's arithmetic: a published template gets copied by thousands of people, lands in the same inboxes, and becomes recognisable within months. The people you're messaging have seen more cold outreach than you have.

So the useful part of any template is the structure underneath it, what each line is doing and why. Read the reasoning under each block, then rewrite the words in your own voice. A mediocre message in your voice beats a polished one in someone else's, because the second one is recognisable and the first one isn't.

The connection request: the note question

"Always personalise your connection request" is the most repeated advice in LinkedIn outreach, and the data behind it is a genuine mess. Rather than pick the study that makes the cleanest story, here is what the published figures actually say.

Source	Acceptance: no note	Acceptance: with note
~20M requests (largest published)	26.37%	26.42%
~64M campaigns	34.3%	32.3%
~16k invitations	Higher	Lower

Two things stand out. The largest dataset finds a difference of five hundredths of a percentage point, which is another way of saying no difference at all. And where a gap does appear, it favours the blank request, modestly, a couple of points, not the transformation the advice implies.

On what happens after acceptance, the sources contradict each other outright. One reports blank requests replying at 20% against 15% for noted ones. Another, on a larger sample, reports the reverse and more emphatically: 9.36% with a note against 5.44% without. Both cannot be right, and neither publishes enough about how their campaigns were targeted for the difference to be resolvable from outside. Treat anyone quoting either number as settled fact, including us, before we checked, with suspicion.

What the disagreement still tells you

Every study that separates generic notes from specific ones finds the same ordering: a generic note performs worse than no note, and a genuinely specific note performs better than either. That ordering is the reusable finding. The headline with-versus-without number is mostly measuring how good each vendor's customers were at writing notes.

Which gives a rule that doesn't depend on whose data you believe: if you can't fill 300 characters with something true only of that person, send it blank and spend the personalization on the first message instead. A blank request is ambiguous, and ambiguity is cheap to accept. A note that could have been sent to five hundred people removes the ambiguity and replaces it with a sales approach that arrived before you earned any attention.

Connection note: mutual connection148 chars

Hi Priya, we both know Alex Grant, he mentioned your team rebuilt onboarding last quarter. Been down that road myself, would be good to connect.

Why it works: The verifiable detail is a person's name, which is expensive to fake and instantly checkable. The last line offers shared experience, not a service.

Connection note: after their post181 chars

Hi Tom, your post on killing the SDR-to-AE handoff stuck with me, particularly the bit about context loss. We tried something similar and got it half right. Happy to compare notes.

Why it works: Quoting a specific argument from the post proves you read it. 'Got it half right' is an admission, which is disarming and almost never appears in cold outreach.

Free accounts: notes are rationed

Free LinkedIn accounts get roughly five personalised invites a month. After that, requests still send, just without the note. A sequence that personalises every invite burns the whole month's allowance in about two days. See LinkedIn connection request limits for the full set of numbers.

The first message after they accept

This is the message that decides everything, and the most common way to lose it is to treat an accepted connection as buying intent. It isn't. They accepted an ambiguous request. You have roughly one screen of a phone to give them a reason to care.

Three jobs, in order:

Prove you looked. One clause that could only be written about them.
Say what you noticed. An observation about their situation, not a description of your product.
Ask something easy. A question they can answer in under ten words, ideally about their own opinion.

First message: the observation opener

Thanks for connecting, Priya.

Noticed you're running outbound across four reps' accounts, most teams that size are still doing it from one and wondering why volume caps out.

Curious how you're splitting it. Territory, or by persona?

Why it works: The middle line describes their situation and implies expertise without claiming any. The question is about their setup, so answering it costs nothing and doesn't commit them to a conversation about buying.

First message: the shared-problem opener

Thanks for the add, Tom.

We spent most of last year trying to get reply rates above 8% on LinkedIn and mostly failed until we cut send volume in half. Counter-intuitive, still true.

Are you seeing the same thing, or is volume still working for you?

Why it works: Leads with a failure, which is rare enough in outreach to read as human. The question invites disagreement, and people reply to disagree far more readily than they reply to agree.

First message: the useful-thing opener

Thanks for connecting, Sam.

We pulled together the actual weekly invite ceilings by LinkedIn account type after getting three accounts restricted learning it the hard way. Happy to send it over if it's useful, no strings, it's a public page.

Why it works: Offers something with a real cost to produce and no conversion mechanism attached. 'No strings, it's a public page' removes the suspicion that saying yes puts them in a funnel.

What not to put in the first message

A calendar link. A deck. The word "synergy". Your funding announcement. A question that is transparently a qualification question ("who handles outbound at {{company}}?"). Anything that makes it obvious the next message will be a pitch regardless of what they say.

Follow-ups: add, don't ask

Most follow-ups fail for the same structural reason. "Just following up" and "bumping this to the top of your inbox" ask the recipient to do the work, to remember your first message, reconstruct why it mattered, and decide. That's a lot to ask of someone who has already implicitly declined once.

A follow-up that works brings something new, so it stands on its own even if the first message was never read.

Follow-up 1: new angle (3 days later)

One thing I forgot to mention: the reason volume caps out isn't usually the tool, it's LinkedIn's weekly invite ceiling. It's per account, so a second sender doubles your ceiling and a bigger daily number doesn't move it at all.

Might be irrelevant to you, figured it was worth 20 seconds either way.

Why it works: Teaches something concrete and free. 'Might be irrelevant to you' gives explicit permission to ignore it, which paradoxically raises reply rates because it removes the pressure to respond.

Follow-up 2: the specific example (5 days later)

Last one from me on this, Priya.

A team about your size was running 4 accounts at 25 invites/day each and hitting the weekly wall by Thursday every week. Dropping to 15/day and adding a fifth sender got them more total volume, not less.

Worth a look at your numbers?

Why it works: 'Last one from me' is honest and it is the line that most often triggers a reply. The example is specific enough to be checkable and describes a mechanism, not a testimonial.

The close-out (7+ days later)

Closing the loop, Tom. I'll stop here.

If outbound volume becomes a problem later, the weekly-ceiling thing is the first place I'd look. Good luck with the quarter either way.

Why it works: A genuine ending with no hook. It reliably outperforms every other follow-up in the sequence, because it's the only one that costs the sender something. Do not follow it with another message; that turns the one honest line in your sequence into a lie.

Touch	Timing	Job
Connection request	Day 0	Get accepted. Blank unless you can be specific
First message	1–2 days after accept	Give them a reason to care
Follow-up 1	+3 days	New angle, teach something
Follow-up 2	+5 days	Specific example or proof
Close-out	+7 days	End it honestly

Five touches over roughly two weeks. If you are tempted to add a sixth, the honest read is that the targeting is wrong rather than the persistence insufficient.

Situational messages

Re-engaging someone who went quiet

Went quiet mid-conversation

Hey Sam, you mentioned Q3 was the realistic window when we last spoke. That's about now, so figured I'd check whether it survived contact with reality.

No pressure if it slipped, most things do.

Why it works: Quotes their own words back, which proves this isn't a mail-merge. 'Most things do' removes the implied criticism of them for the delay.

They changed jobs

New role trigger

Congrats on the move to Northwind, Priya.

New role, so this is probably the wrong month to be thinking about outbound tooling, but it's usually also the one month you get to change how things work before the habits set.

Worth a conversation in a few weeks?

Why it works: Names the objection before they do, then reframes it as the reason to talk. Proposing a later date respects the actual situation and gets a 'yes, later' rather than a silence.

Asking for a referral, not a meeting

Wrong person, right company

Thanks for the reply, Tom, sounds like this isn't your area.

Rather than guess, is there someone on the revenue side who'd actually care about LinkedIn sending limits? Happy to take it off your plate either way.

Why it works: Accepts the no immediately, which is what makes the ask acceptable. People who feel released from a conversation are noticeably more willing to spend a referral on you.

When InMail is the right call

InMail costs credits, arrives in a visually distinct inbox that senior people have been trained to ignore, and carries no relationship context. It earns its place in two situations: the person can't accept invites (a full queue, a closed profile), or you need to reach someone whose invite you'd rather not spend a slot on.

InMail: earn the subject line

Subject: your weekly invite ceiling, not your daily one

Priya, quick technical note rather than a pitch.

If your team's LinkedIn sending has plateaued around 28 invites/day/account no matter what you set, that's the rolling weekly cap, not your tool. It's fixable, but not by raising the daily number.

Happy to explain in three lines if useful. If not, ignore me and the credit's on me.

Why it works: The subject line makes a falsifiable claim about their situation instead of asking for time. Diagnosing a specific number they may be seeing is the strongest opener available, because it proves competence before it asks for anything.

Personalization that actually works

The merge tag is the most over-rated tool in outbound. Recipients have learned to read "Hi {{firstName}}, I see you're at {{company}}" as exactly what it is, because they receive fifteen of them a week and the shape is unmistakable. A first name in the greeting is table stakes, not personalization.

Tier	What it looks like	Reply-rate effect
Merge tags only	Name, company, job title	None, reads as automation
Segment-level	A line true of their role or industry	Modest, better than nothing
Trigger-based	Job change, funding, hiring, a post	Strong, implies a reason to write now
Genuinely specific	Something only true of that person	Strongest, and doesn't scale past ~30/day

The practical resolution: segment tightly enough that a segment-level line is nearly as good as an individual one. Two hundred people who share one specific, painful problem can receive the same message and each feel it was written for them. Two thousand people who share only a job title cannot.

A checklist before you launch

Read the first message out loud. If it sounds like a brochure, rewrite it

First message under 400 characters, connection note under 300

Every merge tag has a fallback that reads naturally when the field is empty

The sequence ends. Somebody who never replies stops hearing from you

Two variants of the first message running against each other, so you learn something

"I hope this message finds you well", it finds nobody well

A calendar link before they've said a single word back

The same template you found on a blog, pasted unedited. Including these

Where the message meets the machine

Good copy and safe sending are the same project. A message that gets ignored lowers your acceptance rate, and a lower acceptance rate tightens the invite ceiling LinkedIn gives your account, so bad copy literally reduces how much you're allowed to send. This is why "send more" and "write better" aren't alternatives.

In Outvelo, each message step can carry variants that split evenly and report reply rate separately, so the question of which opener works is answered by the data rather than by argument. Sequences branch on whether someone replied, so a person who answers stops receiving follow-ups written for someone who didn't. How to structure the sequence covers the mechanics.

Where these numbers come from

Every acceptance and reply figure on this page is published by an outreach platform measuring its own customers' campaigns. Nobody in this category runs a controlled experiment: the populations differ, the targeting differs, and each vendor's number is partly a measure of how good their particular customers are at writing. That is why the studies disagree, and why we show the disagreement rather than averaging it into a single confident figure.

Reply rates vary enormously by market and list quality, treat everything here as direction, not a benchmark you should be hitting. If you have data that contradicts a number on this page, we would genuinely like to see it: [email protected].

LinkedIn Message Templates That Get Replies (2026)