What acceptance rate do B2B LinkedIn connection requests get across seven segments?

Across 300 B2B connection requests I sent from one LinkedIn profile between Aug 10 and Oct 23, 2026, 63 were accepted, a rate of 21.0%. The rate ranged from 6.8% on a cold list of private equity operating partners to 66.0% on founders I had met in person a few days earlier, and the gap between list sources was wider than any gap between industries. The numbers come from my own prospecting campaigns for Growth Marketing Solutions (GMS), pulled from the outreach tool on Oct 26, 2026, and I compare them with three public benchmark pages I opened for this piece. They are the outbound figures behind the European market entry guide.

A connection acceptance rate is the share of connection requests a person accepted, measured at a fixed date after sending. It says nothing about replies, which I report separately below.

What acceptance rate did 300 B2B connection requests get overall?

The 300 requests produced 63 acceptances, 21.0%, measured on Oct 26, 2026. That sits below the public averages, which cluster around a third, and the reason is the mix: two thirds of my requests went to cold database lists, where the rate was about one in ten.

Method, so the numbers are comparable to yours. One personal profile, mine, with a sending cap of around 20 connection requests a day. Every request carried a short note, drafted with AI from research on the person and edited and sent by me; the tools are listed in the rules section below. "Accepted" is the tool's count at pull time, so acceptances that arrived after Oct 26 are not in these figures, and the youngest campaign had only a week to mature.

The whole sample is 300 requests, so the small rows carry little weight. The hotel row is eleven requests and three acceptances, and I do not build anything on it. The rows with 44 or more requests are where I read a pattern.

For comparison, Phantombuster's survey of B2B sales reps, run in Q4 2025, reports that "the average acceptance rate sits between 30% and 40%" (Phantombuster). Cleverly's 2026 benchmark page states that "a real-world study of 16,492 LinkedIn connection requests found an average acceptance rate of 37%" and puts the 2026 benchmark at 30 to 45% (Cleverly). Neither page splits its average by list source, which is the split that explains my 21.0%.

How did acceptance differ between cold lists, job-brief sourced lists and people met in person?

The list source set the rate. Cold database lists accepted at 10.1%, companies that had just posted a brief at 19.6%, and founders I had met in person at 66.0%, about six and a half times the cold rate.

The arithmetic for the cold figure: the five cold segments together are 11 + 44 + 48 + 81 + 15 = 199 requests and 3 + 3 + 6 + 6 + 2 = 20 acceptances, which is 10.1%. Adding the intent-sourced list gives 250 requests and 30 acceptances, 12.0%. The warm list alone is 50 requests and 33 acceptances, 66.0%.

Each step up the table is a stronger reason for the person to recognise the note. A cold database contact has no reason beyond the observation in the note itself. A company that posted a brief for growth help this week has a stated need, and a note that refers to the posting is a reply to something they started. A founder I spoke to on Tuesday is reading a note that says so on Thursday.

Cleverly's page describes the same shape without the data: it says warm outreach after engaging with a prospect's content can push acceptance "above 60%", while cold, context-free requests average 20 to 30% even with solid targeting (Cleverly). My warm rate matches the first claim. My cold rate is lower than the second, and the next section says why I think that is.

Which cold B2B segments accepted most and least?

Among the cold rows with a readable sample, the industrial technology list accepted most at 13.3% and the private equity operating partners least at 6.8%. The German high-LTV B2B list sat at 12.5% and the US startup list at 7.4%.

Two things separate the top and bottom of that range, and I report them as observations from one pull.

The first is research depth per note. The industrial technology list was 15 hand-picked accounts, one campaign per lead, each note written from something specific about that company. Two acceptances is too few to lean on, but it was the most expensive note per person and the highest cold rate. The German list, also researched person by person and written in Sie-form, was second at 12.5% from 48 requests.

The second is who the person is. Private equity operating partners are the most contacted people on my list, and they accepted 3 of 44. Cleverly's seniority table gives cold acceptance for C-level executives at 10 to 20%, against 30 to 45% for managers and individual contributors (Cleverly). My operating partner row sits under the bottom of even that senior band.

I run every cold list at the same daily cap and with the same note rules, so the spread from 6.8% to 13.3% comes from the lists, since send volume was the same for each. Phantombuster's survey makes volume its main finding: reps who send fewer than 25 connection requests a week are "nearly twice as likely to achieve a 40%+ acceptance rate" as those who send 26 or more (Phantombuster). My cap is already below the level where that page sees the drop, and my cold rates are still a third of its average, so for cold database lists I read the list itself as the limiting factor.

How did US startups entering Europe respond compared with DACH companies?

The US startup list accepted at 7.4%, the two German-speaking lists at 12.5% and 27.3%, and the German-speaking notes were written in formal Sie-form. On these samples I cannot say DACH accepts more. The US list accepted at about the rate of the other cold lists.

The US list was 81 requests to startups weighing a European market entry, split across three sub-lists. B2B SaaS founders accepted 2 of 16, 12.5%. DTC and consumer companies accepted 2 of 23, 8.7%. AI and agent companies accepted 2 of 42, 4.8%. The AI list was the largest and the lowest, and those founders are the most pitched people at any tech event this year, so I would not treat 4.8% as a judgement on the note.

The DACH rows are the ones I handle with the most care. The hotel list is eleven requests. The German high-LTV list is 48, and its 12.5% is the number I would quote for a cold, researched, Sie-form note to a German company today. The hub piece on launching in a new European market lays out why LinkedIn is the first channel in Germany, Austria and Switzerland at all: cold email needs prior consent there, and a connection note written as a request to talk does not. The statute detail is in the sibling piece on cold email law in Germany for B2B.

What did acceptances turn into in conversations, replies and interested tags?

Acceptance opened a conversation in every case, because the tool sends the first message after acceptance, and 21 of those 63 conversations drew a reply, 33.3%. On the in-person list, 14 of 33 conversations drew a reply, 42.4%, and 8 of those 33 were auto-tagged as interested.

Definitions from the tool: "conversations started" is its count of first messages sent after acceptance, "replies" is replies to those, and "tagged interested" is its automatic tag on the reply content. The account total over the same period counts 109 conversations and 25 replies, more than the seven segment rows add up to (63 and 21), because it also includes two message-only campaigns to people already in my network: 46 conversations, 4 replies, no connection request involved. That is different work, so the table below uses the seven segments, and its last row counts only conversations that followed an accepted request.

Per 100 requests sent, the gap widens. The five cold lists produced 5 replies from 199 requests (1 + 1 + 2 + 1 + 0), about 2.5 replies per 100 requests. The in-person list produced 14 replies from 50 requests, 28 per 100. On replies the warm list outperformed cold by about eleven times, not six and a half, because the reply rate on conversations was higher as well as the acceptance rate.

Cleverly's benchmark puts a good reply rate on LinkedIn outreach at 10 to 25% and says direct messages to first-degree connections perform at 25 to 35% with personalisation (Cleverly). The 33.3% across the seven segments sits above the first band, and the 42.4% on the in-person list sits above the second; the message-only campaigns to existing connections, 4 replies from 46 conversations, sit well below it.

The interested tags are where the sample gets thin. Ten across 300 requests, eight of them from the 50 people I met in person. I report them as the tool's count; ten is too few to rank segments.

What rules did every connection note follow?

Every note had one specific observation about the person and one question, with no link, no offer and no pricing, and no campaign went past two touches. The rules were the same across all seven segments, so the differences in the table are not differences in note format.

The rules in full:

1. One sender profile, mine, capped at around 20 connection requests a day.
2. One observation from the person's own profile, post, job ad or company news. If the sentence could go to someone else by swapping the name, it did not go out.
3. One question the person can answer in a sentence.
4. No link, no offer, no deck, no "happy to compare notes".
5. German-speaking contacts in formal Sie-form with surname and title where the person uses one.
6. Two-touch ceiling: the connection note, then one follow-up after acceptance if there was no reply. After that the contact rests until there is a new reason to write.
7. A person I had met got the meeting named in the first sentence.
8. AI sourced the accounts, researched the person and drafted the note. I judged, edited and sent every message. HeyReach sequenced the sends, Apollo enriched the accounts, Notion held the pipeline.

Because every request carried a note, I have no control group for the note itself. The one public figure I found on that question is on Cleverly's page, which cites Prospeo data of 26.42% acceptance with a note against 26.37% without (Cleverly). That figure says a note by itself adds nothing. In my data, the difference between 10.1% and 66.0% came from what the note referred to.

On volume, Expandi's September 2026 guide says LinkedIn "doesn't publish a fixed numeric weekly or daily invitation allowance", advises staying within 100 invitations a week at 15 to 20 a day, and notes that an account can be restricted sooner if too many invitations are left pending or ignored (Expandi). A 10% acceptance rate on cold lists means 90 of every 100 invitations sit pending. That alone is reason to keep cold volume low and withdraw old invitations.

What should a founder or growth lead do with these numbers?

Spend the week before outbound on list quality and on reasons to write; note copy moved less than either in this pull.

What I would do, in order:

1. Rank every source of contacts by the reason it gives you to write. Met in person, referred by name, posted a brief or job this month, engaged with your content, cold. Fill the week from the top of that list before touching the bottom.
2. Put one event a quarter in the target market on the calendar and go with a list of people to meet. Fifty requests to founders I had met during one week in San Francisco produced 33 connections and 14 replies; no cold list in this pull came close per request sent.
3. Watch postings. A company that has just asked a freelance platform for growth help accepted at one in five, and the note writes itself from the posting.
4. Keep cold database lists short and researched. Budget the research time per name, and if there is no time, cut the list rather than the research.
5. Measure acceptance at 30 days with the pull date written next to the number, and measure replies on conversations separately. A rate quoted without its measurement date and list source is not comparable to anyone else's.
6. Treat the public averages as a mixed-source figure. A cold list that accepts at 10 to 13% with a researched note is in line with what I see.

The segment to pick first for a European entry follows from the first-channel decision: one channel, one segment, enough volume to read a rate. On LinkedIn, enough volume is 50 requests per segment before you compare.

This is the outbound work GMS runs for companies entering a European market, either as a Fractional Head of Growth engagement of 3 to 6 months at about two days a week, or as the Pipeline Build, 6 weeks, priced on request, which takes the GTM Sprint output and builds the account lists, LinkedIn sequences, founder content, demo booking, CRM hygiene and handover. Both are fixed fees scoped after a call, never hourly.