LinkedIn Reply Rate 2026: Benchmarks and Real Numbers

In short: In 2026, a typical LinkedIn campaign gets roughly one acceptance for every five connection requests and a reply from roughly one in five accepted contacts. That is what HeyReach's data shows. Expandi reports 28.5% acceptance and 10.4% replies, but counts replies per message instead of per conversation. In two CompLeadly accounts, the acceptance rate per audience ranged from 12.6% to 36.7%. Only compare your rate with numbers that were measured the same way, and measure cleanly first.

As of: October 2026.

What counts as a good LinkedIn reply rate depends first on what you divide by what. Two well-known tool vendors published benchmarks in 2026, and one reply rate is more than twice the other. Both calculate correctly. They just calculate differently.

This article puts the numbers in context. You get the figures from Expandi and HeyReach with their methodology, our own acceptance rates from two CompLeadly accounts, evaluated in October 2026, a guide to measuring your own rate, and a calculation of what a low rate costs you in working time.

Three rates that often get mixed up

When people talk about "reply rate", they mean one of three numbers. Each has a different denominator.

The acceptance rate measures the first step. A person sees your profile and your request and decides whether to let you into their network. If it is low, the audience usually doesn't fit, or your profile doesn't say clearly who you work for.

The reply rate measures the conversation after that. It tells you whether your first message gives people a reason to reply. What matters is whether you count per contact or per message. If you send three follow-ups to one contact and get one reply, that is 100% per contact and 25% per message.

Replies per request is the number that counts for sales in the end. It combines both: 25% acceptance and 20% replies give you 5 replies per 100 requests.

What Expandi and HeyReach measure in 2026

Both companies sell software for LinkedIn outreach and evaluate campaigns that ran on their own platform. That is a large data set, but not a neutral one: it measures what customers of these tools do. How many people send their requests by hand, and with what result, is not part of either report.

HeyReach: 96,051 campaigns, measured per conversation

In September 2026, HeyReach published data from 96,051 campaigns over the previous twelve months. Each campaign counts on its own, not as the average of a workspace (HeyReach, LinkedIn outreach benchmarks 2025 to 2026, accessed on 12 October 2026).

Two findings stand out. 10.7% of the campaigns that got acceptances received not a single reply. So the loss often happens after acceptance, not before. And the acceptance rate drops only slightly as volume grows, from 21.43% at 50 to 100 requests to 19.35% at 1,000 requests and more, while the reply rate stays almost the same.

Expandi: 13.2 million requests, measured per message

Expandi evaluated 13,218,869 connection requests and 6,730,447 outbound messages from 13,302 accounts, covering May 2025 to April 2026 (Expandi, LinkedIn Outreach Benchmarks 2026, accessed on 12 October 2026).

Under "all outbound messages", Expandi counts welcome messages, follow-ups, group and event messages, InMails and emails. So every follow-up increases the denominator. The acceptance rate stayed between 28% and 30% in eleven of twelve months, while replies to notes fell from 3.5% to 2.2%.

Expandi also breaks the numbers down by industry, namely the industry of the sender, not of the person contacted:

Why 10.4% and 22% don't contradict each other

HeyReach divides replies by started conversations, Expandi by all sent messages. If you send each contact an opener and three follow-ups, Expandi's calculation puts up to four times as many messages in the denominator as there are conversations. The number gets smaller, although just as many people reply. On top of that, HeyReach gives the median of individual campaigns, Expandi an overall value across all requests and messages.

For you, that means: use the number whose calculation matches your own. If you count per contact, compare yourself with HeyReach. If you count every message, compare yourself with Expandi.

Our numbers: acceptance rates from two CompLeadly accounts

In October 2026 we evaluated the acceptance rate per audience in two CompLeadly accounts. We counted requests sent with the "Visit + Connect" feature. The requests are from 2025 and 2026.

We take three things from this.

The range is wide. Between 12.6% and 36.7% lies almost a factor of three. A single industry value tells you little about your own audience.

The same audience varies by region. Coaches in North Rhine-Westphalia accepted at 31.8%, coaches in Baden-Württemberg and Berlin at 12.6%. Both audiences ran in the same account. So testing pays off at the level of the audience, not just the message.

Small lists are not benchmarks. 40 requests to the UK or 68 to real estate decision-makers show a direction, not a reliable rate. The more requests behind a rate, the more you can rely on it.

We don't have a reply rate of our own. The evaluation this table comes from contains requests and acceptances, but no reply rate. Prospects book meetings with us through an external booking link, so the meeting rate is missing as a counted number too. Where we calculate with meetings below, it is our own estimate and marked as such.

How to measure your own reply rate

A benchmark only helps if your own number is calculated the same way. Here is how to measure cleanly.

1. Build cohorts. Treat all requests to one audience in one week as a group. Evaluate it only after three weeks, so late acceptances and replies count too.
2. Count people, not messages. A contact who replied counts once, whether they replied to the opener or to the third follow-up. Then your number matches HeyReach.
3. Separate acceptance and reply. Calculate both rates separately, plus replies per request. That is the only way to see where the process gets stuck.
4. Count only real replies. A "thanks for connecting" without content is a reply, but not a conversation. Track it in a separate column if you want to tell them apart.
5. One audience, one message per cohort. If you change message and audience in the middle of the week, you won't know afterwards what worked.

Be careful with tools that calculate acceptances from contact growth. If someone accepts a request from last month, or sends you a request themselves, it counts as a new contact this week. CompLeadly calculates its acceptance rate this way too, from contact growth divided by sent connection requests, over the last seven days in the dashboard. That is why the value can exceed 100%. It is fine for a quick look at the week; for a clean comparison with a benchmark, use the cohort.

Our recommendation: From now on, measure per contact and per audience, in one-week cohorts. Compare yourself with your own previous weeks first and only then with HeyReach. Expandi's message rate is only a useful comparison if you count every message yourself.

Why your rate is low

HeyReach offers a simple diagnosis: if acceptance rate and reply rate are both low, the problem is usually the audience. If acceptance is good but the reply rate is low, it is the first message (HeyReach).

Acceptance rate below about 13%: Check the list first. Who is in your audience that you don't actually want to reach? A sharper ideal customer profile does more here than a new message. Then comes your profile: headline and About section have to say who you work for. Concrete levers are in Increase your LinkedIn acceptance rate, the question of the note in LinkedIn connections with or without a message.

Good acceptance, hardly any replies: The first message after acceptance carries the weight. It needs a reference to the person and a question that is easy to answer, and no offer in the first sentence. Examples are in Thanks for connecting: messages after acceptance and in the connection message templates.

Replies only come after a follow-up: Plan at least one follow-up with new content instead of a plain reminder. Wording and timing are in Follow-up templates with timing.

Lots of pending requests: Count requests that are still pending after three weeks as not accepted. Then withdraw them to make room for new requests, as described in LinkedIn pending connections.

InMails follow their own rules, more on that in How to get replies to your LinkedIn InMails.

By hand: 100 requests and their rates

This is the work behind a cohort of 100 requests if you click everything yourself. The minutes come from two CompLeadly accounts in 2026. They are the measured execution time of the software per action and therefore a lower limit for manual work: the software pauses like a human when opening, reading and typing.

1. Define the audience in LinkedIn search or Sales Navigator: not measured.
2. Open and follow 100 profiles: about 150 minutes.
3. Read 100 profiles and send the connection request: about 300 minutes.
4. After a week, check who accepted and enter it in a spreadsheet: not measured.
5. Write the first message to about 25 new contacts: about 38 minutes.
6. About 25 follow-ups to contacts without a reply: about 30 minutes.
7. Withdraw pending requests after three weeks: at most about 20 minutes (calculated for all 100).
8. Count replies, calculate rates, handle conversations: your work, not measured.

That is about nine hours of clicking for a single cohort, not counting the counting.

How it works with CompLeadly

CompLeadly is a desktop app that runs on your computer. It takes over steps 2, 3, 5, 6 and 7. For steps 4 and 8, it provides the numbers you need from Premium.

• Audience (you, once): A LinkedIn people search, a Sales Navigator search or an Excel list becomes an audience. Create a separate audience for each region and each message test, so you compare rates per audience instead of an average.
• Connecting, opener, follow-up (software): Profile visit and connection request. After acceptance a first message goes out, but only into an empty chat. A follow-up only comes once your last message has gone unanswered for as many days as you set. If the contact replies, the automatic messages stop.
• Withdrawing (software): CompLeadly withdraws pending requests that have gone unanswered for too long.
• Numbers (software, from Premium): Real-time KPIs, a replies overview of who replied, and a weekly report. The report shows the acceptance rate and breaks down acceptances and replies by feature, audience and message. How the replies table works is explained in the knowledge article How does the "Replies" table work?
• Replies (you): You hold the conversation yourself. With the separately bookable Smart Chat, your LinkedIn messages sit in CompLeadly, you filter for unanswered chats and assign chats to colleagues, who don't need access to your LinkedIn profile for that.

Connecting, opener and follow-up already run in the free version, with daily limits such as 2 connection requests, 10 messages to 1st-degree contacts and 1 follow-up per day. Unlimited and as a sequence, they come with Basic at $29.99 per month excl. VAT, or $24.99 on the yearly plan. The analytics are part of Premium at $59.99 per month excl. VAT or $49.99 on the yearly plan. You start with a 10-day trial, no credit card. All packages are on the pricing page, as of 12 October 2026. How CompLeadly differs from Expandi, HeyReach and other tools is shown in the tool comparison.

Time calculation

The table calculates the clicking for 100 contacted people, up to the booked meeting.

How strongly the acceptance rate drives manual work is shown by the same calculation with our lowest and highest rate:

At 12.6%, each new contact costs you over two and a half times as much clicking time by hand as at 36.7%. A better audience saves more time than any clicking trick. With CompLeadly the clicking disappears, but the rate is still your most important lever, because it decides how many conversations your audience produces.

Recruiting and job hunting: different rates, same calculation

Recruiting: In Expandi's data, senders from Staffing & Recruiting reach 36.5% acceptance and an 18.9% message reply rate, both clearly above the platform's overall value (Expandi). As a recruiter, compare yourself with this row, not with 28.5%. With your own Recruiter Lite license, CompLeadly turns a Recruiter search, a pipeline or a saved audience into an audience, and connecting, opener and follow-ups run until someone replies. Wording is in LinkedIn recruiting messages, the process on the page for recruiters.

Job hunting: As a candidate you don't message a hundred people, but twenty hiring managers and recruiters at your target companies. With numbers this small, every rate is chance. Keep count anyway, but per company: who accepted, who replied, where is a second message worth it? In CompLeadly, a people search for these roles becomes an audience, and profile visit, connecting, opener and follow-up run in the free version within its daily limits. For twenty target companies that is enough. What to avoid along the way is in LinkedIn mistakes when job hunting.

In sales, the same calculation as above applies. How replies turn into a predictable flow of meetings is shown in Customer acquisition: the process by hand and automated, the metrics in software sales in SaaS sales on LinkedIn.