A good reply rate for LinkedIn outreach is one that beats the published average for the same denominator: above 10.4% of messages sent, above 18.1% of accepted connections, or above 7.2% across every contact in the campaign, counting every reply including the noes (Expandi, HeyReach, Belkins). The question "what is a good reply rate for LinkedIn outreach" gets answers from single digits to above half because each answer divides by a different number. Pick the denominator first, then the benchmark.
Why does a good reply rate for LinkedIn outreach matter?
Reply rate matters more on LinkedIn than on email because LinkedIn caps how many conversations you can start. Intentsly's own safety limits are 30 messages and 30 connection requests per account per day, with 15 recommended, and up to 2 accounts per plan. An email tool can send a thousand more messages tomorrow; a LinkedIn account cannot, so the only way to get more replies is a higher rate on the same small number.
Take a labelled hypothetical. Halden Payroll Software sends 300 connection requests in a month to finance leads at 50-person firms. Seventy-five accept and 15 of those reply. The founder reads a vendor page that says a good reply rate is one in five, and another that says one in ten, and cannot tell whether 15 replies is good or bad. The answer is that it is one in five of accepted connections and one in twenty of requests sent, and both figures sit close to the published averages, as the next section shows.
How is a LinkedIn reply rate calculated?
A LinkedIn reply rate is replies divided by something, and the three somethings in common use give three different numbers from the same campaign. The large 2026 datasets each use a different one, which is why their headline figures look unrelated.
| Denominator | Formula | Who publishes it this way | 2026 figure |
|---|---|---|---|
| Messages sent | Replies to outbound messages / outbound messages sent, follow-ups included | Expandi, 6,730,447 messages, 13,302 accounts, May 2025 to April 2026 | 10.4% |
| Accepted connections | Replies / accepted connections, per campaign | HeyReach, 96,051 campaigns, 12 months to September 2026 | 18.10% typical, 21.23% average |
| Contacts in the campaign | Replies across everyone who entered, connected or not; the study's methodology line says replies divided by messages sent, so the exact denominator is not fully clear | Belkins, 15 million-plus contacts, 2025 | 7.2% |
Apply the worksheet to Halden. Requests sent 300, accepted 75, outbound messages sent 180 (two or three per accepted contact), replies 15. Replies per contact: 15 of 300, one in twenty, a little under Belkins' 7.2% headline. Replies per accepted connection: 15 of 75, one in five, in line with HeyReach's 18.10% typical campaign. Replies per message: 15 of 180, about one in twelve, in line with Expandi's 10.4% (Expandi, HeyReach, Belkins). Same campaign, three correct rates, and Halden is about average on all of them.
A reply that counts in every benchmark but books nothing looks like this:
Thanks for reaching out. Payroll is handled by our accountant and we're not looking to change this year.
What is a good reply rate for LinkedIn outreach in 2026?
Good means the upper band of the dataset that uses your denominator. Only one of the three sources publishes bands; the other two give an average and a few comparison points.
| Denominator | Weak | Typical | Strong | Source and limits |
|---|---|---|---|---|
| Accepted connections | Below 9.62% | 18.10% | Above 28.89% | HeyReach, per-campaign bands from 96,051 campaigns; one vendor's customers; no exact calendar dates |
| Connection acceptance (for context) | Below 12.97% | 20.75% | Above 31.78% | HeyReach, same report |
| Messages sent | No band given | 10.4% platform average | Staffing and recruiting senders reach 18.9%; computer software senders sit at 8.8% | Expandi, sender-side data, all replies counted, no positive split |
| Contacts in the campaign | No band given | 7.2% overall | The study calls anything above 8% in a US campaign excellent | Belkins, updated 29 June 2026; also reports 18.7% acceptance and 1.3% of connected prospects booking a meeting across 14,077 contacts in 34 client projects |
Two more findings from the same sources change what "good" should mean for you. HeyReach reports that 10.7% of campaigns with accepted connections got no reply at all, so a campaign with any replies is already ahead of one in ten (HeyReach). Expandi's sequence data, 2,966 campaigns, puts three-message campaigns at 9.8% replies against 6.3% for one message and 5.0% for five or more, so a rate measured on a one-message campaign is not comparable to one measured on a three-message campaign (Expandi H2 2026).
One first-hand figure, with its limits. When Intentsly launched we ran our own signals and AI SDR on LinkedIn, targeting founders who had posted about cold outreach burnout, and the reply rate was 4x our previous cold campaigns, with 14 qualified demos booked in 14 days (case study). That is one company over two weeks, and the people entered the outreach because they had posted about the problem that week, which is the single biggest reason the rate moved; a warm entry rule does more than any wording change.
Reply rate vs acceptance rate vs positive reply rate
The three terms get mixed up on most benchmark pages. Keep them apart and most "my rate is low" questions answer themselves.
| Term | What it measures | What a low number usually means |
|---|---|---|
| Connection acceptance rate | Accepted / requests sent | The list or the profile: wrong people, or a profile that does not explain why you are connecting |
| Reply rate | Any human reply / your chosen denominator | The message, if acceptance is healthy; the list, if acceptance is also low |
| Positive reply rate | Replies that ask a question or agree to a step / same denominator | The offer or the timing: people answer, but the thing you ask for is not wanted this month |
The diagnostic in the middle column is HeyReach's: low replies with healthy acceptance is likely a messaging problem, low replies with low acceptance is likely a targeting problem (HeyReach). The lesson on what a warm lead is covers the entry-rule side of the targeting problem.
What actually moves a LinkedIn reply rate?
The levers with data behind them are entry (who gets in), sequence length and a note on the request. Wording and length matter less than the pages selling templates suggest.
Belkins' split of 11.5 million connection requests found that a personalised note lowered acceptance slightly, 25.3% against 27.6%, and raised replies from 5.3% to 8.2% (Belkins). Expandi's 458 single-message campaigns put messages of 151 to 200 characters at 8.5% replies and messages over 500 characters at 5.8%, and its within-account test of 118 campaigns found human-written requests accepted about 12% more often than AI-written ones, with reply rates effectively identical (Expandi H2 2026). LinkedIn's own study of tens of millions of recruiter InMails, May 2021 to April 2022, found messages under 400 characters answered 22% more often than average and messages over 1,200 characters 11% less, with individually sent messages about 15% ahead of bulk sends; that is recruiting, not sales, so read it as direction (LinkedIn).
The entry rule is the lever none of the datasets isolates, because every one of them pools cold lists with warm ones. If the person you message posted about the problem this week, the first line can quote them, and nobody in a cold list gets that. Finding those people by hand means reading posts every morning; Intentsly watches LinkedIn keyword posts, comments and competitor engagement and hands you that day's prospects scored against your ideal customer profile, so the entry rule is applied before you write a word.
How to use this lesson this week
Three actions, each under an hour.
- Fill the worksheet for your last campaign. Requests sent, accepted, messages sent, replies, positive replies. Compute replies per accepted connection and replies per message, and write the denominator next to each number so nobody on the team compares a per-accepted figure with a per-message one again.
- Run the two-question diagnostic. Acceptance under one in eight and replies under one in ten of accepted: rebuild the list before touching the copy. Acceptance healthy and replies low: rewrite the first message to under 200 characters and add one thing you know about the person. The guide to warming up a LinkedIn account covers the case where acceptance is low because the profile is new.
- Add a positive-reply column and a date. Count only replies that ask a question or agree to a step, and record the week. Rates drift with the season and with how long the campaign has run, so a number without a date is not a benchmark.
Related lessons
What is cold outreach? defines the first message when there is no prior contact. What is a sales cadence? covers the sequence length that the Expandi data says matters. What is an AI SDR? defines the kind of outreach the 4x figure above comes from. The Learn index lists the rest.
The fastest way to raise the rate is to change who enters the campaign, and the free Buying Signal Playbook lists 12 buying signals worth chasing and 8 to ignore, each with the LinkedIn trigger that reveals it and a script, so your next campaign starts with people who already raised the subject.