LinkedIn Outreach Benchmarks 2026: The One Number Nobody Publishes
TL;DR: The 2026 LinkedIn outreach benchmarks that hold up are 28.5% connection acceptance, 3.0% connection-request reply, and 10.4% post-connection message reply, from Expandi’s 13.2 million request dataset. Run those three through 100 requests and you get about three replies. Not three meetings. Three replies, several of which say no. Every published benchmark stops one step before the number you actually need.
That gap is the whole story, so start with the numbers that are real.
The 2026 LinkedIn outreach benchmarks worth quoting
There is one dataset in this category big enough and specific enough to cite. Expandi’s 2026 outreach benchmarks cover 13,218,869 connection requests, 3,766,161 accepted connections, and 6,730,447 outbound messages, sent from 13,302 accounts between May 1, 2025 and April 30, 2026.
- Connection acceptance: 28.5%
- Connection-request reply: 3.0%
- Post-connection message reply: 10.4%
- Industry spread on acceptance: 17.5% to 40.1%
(Methodology note: vendor aggregate, measured sender-side across mixed industries, and the senders are all people who bought outreach software. Directional. It tells you what the tooling-heavy end of LinkedIn looks like, not what LinkedIn looks like.)
Those four lines are the honest version of what most “LinkedIn outreach benchmarks 2026” posts stretch across three thousand words. Everything past this point is what the numbers mean, which nobody bothers with.
Why the benchmarks don’t stack into a funnel
Here is the first trap. Those rates are measured against different denominators, so you cannot chain them.
Connection-request reply rate of 3.0% is replies divided by requests sent. Message reply rate of 10.4% is replies divided by messages sent to people who already accepted. Different populations, different steps, and the second one has survivorship baked in, because the people who accepted are by definition warmer than the people who did not.
So when you see a benchmark table listing 28.5% and 10.4% side by side, they are not two steps of one funnel. They are two separate measurements that happen to share a chart. If you compare your own end-to-end rate to either one, you are comparing your whole funnel to somebody else’s single step, and you will conclude you are failing when you are just counting differently. This is the same problem I picked apart in social selling statistics, where most of the widely quoted figures turn out to have no published methodology at all.
The number nobody publishes
Do the arithmetic anyway, carefully, and label it as arithmetic rather than as a benchmark.
Send 100 connection requests. At 28.5%, about 29 people accept. Message all 29. At a 10.4% message reply rate, roughly three people reply.
Three replies per 100 requests. And a reply is any response, including “not interested,” “who is this,” and the polite version of both. Nowhere in Expandi’s report, or in any of the vendor benchmark reports I checked for this post, is there a published meetings-booked rate or a closed-revenue rate against outreach volume. The funnel stops at reply, every time.
That absence is not an oversight. Reply rate is the last metric that still looks good. The step after it is where cold outreach math turns ugly, and no outreach vendor has an incentive to print it. So the single most useful benchmark for anyone deciding whether to run LinkedIn outreach at all, meetings per hundred touches, is the one you have to compute yourself from your own numbers.
The trend line matters more than the average
Buried in the same dataset is the finding that should change what you do this quarter. Connection-request reply rates fell from 3.5% in May 2025 to 2.2% in April 2026, a 37% relative decline in twelve months. Post-connection message reply rates held roughly flat over the same window.
Cold is decaying. Warm is not. The front door where you arrive as a stranger is getting worse at a measurable rate, and the door where the person already accepted you is holding steady.
The other piece worth knowing: Belkins’ LinkedIn outreach study, across 14,077 tracked contact records, found that attaching a personalized note to a connection request lifted replies from 5.3% to 8.2% while slightly lowering acceptance. The note pays off after the accept, not before it. I went through that split in detail in LinkedIn connection request acceptance rate.
One benchmark I will not give you is InMail. Search for an InMail response rate and you will find figures from under 5% to over 40% presented as averages, sometimes within the same article. Numbers that far apart are not measuring the same thing, and none of the ones I turned up trace back to a dataset anywhere near the size of Expandi’s. Do not benchmark against a number you cannot source.
Why I stopped comparing my numbers to anyone else’s
I used to check my acceptance rate weekly, the way you check a step count. It went up whenever I got less picky about who I sent to, which felt like progress and was the opposite. More accepts, no replies, and a network full of strangers I could not use to find coaching clients.
What fixed it was changing the order, not the copy. I stopped starting with a list and started with the feed. Find people already posting about the problem I coach on, leave a comment worth reading, wait a day, then send the request. My acceptance went up, but the part that mattered is that people wrote back, because by then I was not a cold name in a notification.
The cost was time. Finding those people meant most of an hour of scrolling before I had done any real work, and on a heavy client day it did not happen at all. That specific chore is why I built LeadBase. It surfaces who engaged with what, so the warm names are already queued when I sit down. I still write every message myself.
Build your own benchmark in 20 minutes
Open a spreadsheet and write down four numbers from your last 30 days, each with its denominator spelled out next to it:
- Requests sent. Count withdrawn and still-pending invites as not accepted. Anything else flatters the number.
- Accepted / sent. This is your acceptance rate. Compare it to your own last month, not to 28.5%.
- Replied / messaged after accept. Your real reply rate, on the same denominator Expandi uses.
- Calls booked / requests sent. The number no vendor publishes. This is the only one that tells you whether the channel pays.
Run it monthly. After three months your own trend line is worth more than any industry average, because it controls for your vertical, your offer, and your face. The tactical follow-on, changing who you message rather than what you write, is in how to improve LinkedIn reply rates.
FAQ
What is a good LinkedIn outreach acceptance rate in 2026? The platform-wide average is 28.5% across Expandi’s 13.2 million request dataset. Roughly 30 to 45% is healthy. Below 20% points at targeting, not wording, since industry moved acceptance from 17.5% to 40.1% in the same data.
What is a good LinkedIn message reply rate? 10.4% is the 2026 average for messages sent to people who already accepted your connection request. Measure yours on that same denominator or the comparison is meaningless.
How many meetings should 100 LinkedIn connection requests produce? No published benchmark answers this, which is the point. Chaining the published averages gives about three replies per 100 requests, and only some of those replies become calls. Track your own calls-booked over requests-sent for 30 days.
Are LinkedIn outreach benchmarks getting worse? The cold half is. Connection-request reply rates dropped from 3.5% to 2.2% between May 2025 and April 2026. Post-connection message reply rates stayed flat over the same year.
Should I trust InMail response rate benchmarks? Be careful. Published figures swing from under 5% to over 40%, a spread too wide for them all to be measuring the same thing. Treat any benchmark without a stated sample size and date range as marketing copy.
What to do next
See how it works. LeadBase surfaces who engaged with what on LinkedIn, so there is a warm name in the queue before you spend a request. Take a look.
Go deeper. If your reply rate is the number you want to move, start with who you are messaging rather than the message itself: how to improve LinkedIn reply rates.
Start free. Stop benchmarking against strangers and go build your own baseline. Start your free LeadBase trial.
