On this page (8)
The three numbers to hold yourself to
28.5%
of connection requests are accepted
Source: Expandi, H2 2026
10.4%
of messages sent after connecting get a reply
Source: Expandi, H2 2026
3.0%
of connection-request notes get a response
Source: Expandi, H2 2026
Valley's 2026 benchmarks add a useful way to think about the spread. Cold template outreach replies at 1 to 3%. Personalized cold outreach lifts that to 5 to 10%. Warm outreach built on a real signal runs 15 to 45%. Same channel, completely different results, and the difference is almost never the tool.
Reading your acceptance rate
Acceptance tells you whether strangers trust you enough to let you in. It depends on two things: who you ask, and who's asking.
Industry matters a lot. In Expandi's data, acceptance ranged from 17.5% in consumer electronics to 40.1% in broadcast media. So compare yourself to your own market, not a blog post's headline.
Who's asking matters less than people think. C-level senders got 29.4% acceptance, junior senders 26.3%. What did move it: accounts with 1,000+ followers and consistent engagement got 30.5%. A profile people recognize gets opened.
Reading your reply rate
Reply rate is where the copy and the sequence show up. Two findings from the report are worth taping to your monitor.
Length. The best band was 151 to 200 characters, at an 8.5% reply rate. Messages of 301 to 500 characters dropped to 7.6%. That's two sentences, not a paragraph.
Sequence length. More touches help, until they don't:
- 1 message: 6.3% reply rate
- 2 messages: 6.8%
- 3 messages: 9.8%, the peak
- 4 messages: 7.9%
- 5 or more: 5.0%
The second message converted best per recipient, at 8.8%. So the follow-up isn't a formality. It's your best shot. Make it a new angle, not a bump.
One more thought, Tom. Most teams your size lose a full day a week to manual follow-ups. If that's you, I can show you how we cut it to zero. Worth a look?
Hi Tom, just bumping this to the top of your inbox in case you missed it!
Timing: what matters and what doesn't
Time of day matters for acceptance. Requests sent between 7 and 11am were accepted at 32 to 37%. Evening sends got 23 to 24%. People triage invitations over their morning coffee.
Reply rate, on the other hand, stayed flat across the day. And the day of the week barely mattered at all: acceptance ranged from 26.5% to 30.4% and replies from 8.3% to 9.1%. Stop agonizing over Tuesday versus Thursday. Send requests in the morning, and send the rest whenever it's ready.
Why last year's playbook scores lower now
One number in the report should change how you work. The reply rate on connection-request notes fell from 3.5% in May 2025 to 2.2% in April 2026. That's a 37% drop in a single year.
-37%
relative drop in replies to connection-request notes, May 2025 to April 2026
Source: Expandi, H2 2026
The likely reason is simple: everyone started pitching in the note, a lot of it written by AI, and people stopped reading notes. So if you benchmark against an old playbook that says “personalize the note and pitch there,” you'll think your campaign is broken when the channel just moved.
The fix is to move the effort. Keep the request plain, spend your research on the first message after acceptance, and judge the campaign on replies to messages, not replies to notes. That's where the 10.4% lives.
A 10-minute diagnosis
- 1Acceptance well under your market's norm? Tighten the list and fix the profile photo, headline and banner.
- 2Good acceptance, replies under 5%? Your first message is pitching or rambling. Cut it to 200 characters and one question.
- 3Replies only on message one? You're not following up, or your follow-ups repeat the ask. Add a second and third message with new angles.
- 4Five or more messages in the sequence? Cut back to three. Past that, you're mostly training people to ignore you.
- 5Plenty of replies, few meetings? The problem is after the reply: slow answers, pitching too early, or no easy way to book.
What it does on LinkedIn, step by step.
Bazooka tracks these numbers for you on every campaign: acceptance, replies, positive replies and booked calls, down to which opener and which follow-up produced them. It collects leads from Sales Navigator searches, post commenters, group members and your own followers, researches each one, and reaches them with an InMail or a connection request after a profile view.
Then it acts on the numbers. Every message, reply and booked call feeds back in, the system keeps what works and drops what doesn't, and the AI closer turns replies into meetings on your calendar.
Questions people ask
What's a good LinkedIn acceptance rate in 2026?
The average across Expandi's 13,302 accounts is 28.5%, but industries range from 17.5% to 40.1%. Aim for your own market's average first, then beat it.
How many follow-ups should a LinkedIn sequence have?
Three messages performed best in Expandi's 2026 data, at a 9.8% reply rate. Five or more dropped to 5.0%.
How long should a LinkedIn message be?
151 to 200 characters had the best reply rate in the 2026 data. Two short sentences and a question is usually enough.
Sources
Get the next playbook in your inbox.
Leave your email and we will send you new guides on LinkedIn and the other channels we run.