Cold LinkedIn benchmarks in 2026, and what actually moves them

The three numbers that matter, in order

Cold LinkedIn has exactly three gates, and they fail in a fixed order.

Accept rate. The share of connection invites that turn into a connection. This is a judgment made in two seconds on your photo, your headline and the mutual context. Our fleet reading is 24.6%: 192 accepts on the 782 invitations sent between 10 and 23 September, each with at least seven days to be answered. Seven days is enough to settle the question, because 183 of those 192 accepts landed inside the first week.

The spread between senders is the part worth staring at. Across the ten profiles that sent 30 invitations or more in that window, accept rate runs from 13.0% to 42.4%. That is a factor of three between real people writing to their own markets, and the two profiles at the top of the range are the ones whose owner works in the industry they write to. The profile decides this gate, and whose name is on the sender moves it further than any rewrite of the message.

Reply rate. The share of people who received a first direct message and wrote something back. This is where copy lives. The published benchmark for cold LinkedIn sits at 14.1%. Our own reading is 18.4%: 70 replies from the 380 people whose first direct message went out between 30 August and 22 September. The same query on the previous window read 20.3% on 187 people, so the rate held its shape while the base doubled. One number without the other is marketing; the pair is a measurement.

Meeting rate. The share of replies that turn into a booked call. Earlier versions of this article left this gate unmeasured, because the count was too small to print. It is now large enough to be worth writing down and still small enough to be honest: of the 191 people who have ever replied to us, 9 booked a call and 8 more qualified as people worth one. That is 4.7% of repliers on the calendar. This gate is decided by what you do in the two messages after the first reply, and the reply handling system is the part of the machine that owns it.

A pipeline that reports only the last number is hiding where it leaks. A pipeline that reports only the first is measuring how likable the profile is.

What moved ours

1. A reason to write, not a list to write to

The difference between 5% and 25% replies is rarely the sentence structure. It is whether the person has a live reason to hear from you this week. We queue people on signals: a role opened three days ago on the company's own job board, a funding filing that named the executives, a post where someone described the exact problem out loud, a technology detection showing what they run.

The test is simple. If the first line of the message would still be true a year from now, it is a description rather than a signal.

2. One checkable fact per message

Every message we send carries a fact with a source behind it: a specific job posting, a specific post, a specific product change. A lead without a fact is dropped from the batch rather than written to with a generic opener. This rule costs volume and buys the only thing that matters at the top of the funnel, which is the reader believing the message was written for them.

The cost is real: on some signals a third of a sourced list fails the fact check. That third produces the replies that make people say outbound is dead.

3. Saying what you sell in the first message

We spent a week measuring a four-message thread that asked four thoughtful questions and never once said what we do. The person answered the first two, shortened the third answer to three words, and went quiet. Five days of a stranger's attention spent, and they still could not have told a colleague what we sell.

Now the first touch says the product in one line, the message stays under 400 characters, and by the third message from us there is a proposed time. Diagnostic questions are excellent in a conversation with a buyer who already knows what you sell. They are expensive with someone who is hearing from you for the first time.

4. Pacing that looks like a person

Volume is the easy dial. The account is the limit that decides how far you can turn it. Sends spread across the whole working window, one cold touch per profile per minute at most, a warm-up ramp for a new sender, and a step back the moment the platform pushes back.

We learned the last part the expensive way: eight profiles out of thirteen spent five days retrying invites into a temporary limit, three times a day. The rule we run now costs one attempt rather than one day. A refusal parks that profile for half an hour and sends it to a different person, and three refusals in a row with nothing sent park it until the next working day.

That leaves a health reading worth watching every week. Across the five weeks to 28 September the platform refused 76 of 2 231 invite attempts, which is 3.4%, and the worst single week was 7.0%. When this climbs, the answer lives in pacing and audience quality, and it shows up in the accept rate about a week later.

What the numbers look like on a normal week

Take the five weeks from 24 August to 20 September and keep the weeks where a profile sent at least 20 invitations. That is 33 sender-weeks. The average one is 43 invitations, 11 accepts and 2 replies, with the weekly invite count running from 24 to 60.

Multiply by senders, not by hope. The engine is linear in profiles, and the growth people expect from outbound comes from adding good senders, better signals, and a higher accept rate on the profiles you already run.

Meetings run on a slower clock than a week. Nine booked calls across the whole life of the engine means you plan capacity in senders and months. A meeting a week from a single cold profile describes a good week rather than an average one.

Two things distort this shape. A tight signal can double the accept rate, and a bad list can halve it while everything else stays identical. Both are visible in a week if every batch carries its signal into the database, which is the entire reason we tag batches at all.

How to read your own numbers honestly

These two rates are recomputed from the engine database and kept current on the benchmarks page, which carries the date of its last recompute at the top. This article is the reasoning; that page is the live reading.

Three rules we hold ourselves to.

1. No decision on noise. Under 30 sends or under 3 replies, a difference between two variants is not a difference. Wait for the sample, and read how to A/B test on small samples before calling a winner.
2. One change at a time. Two changes in the same week make the week unreadable. It feels slow and it is the only thing that compounds.
3. Every number carries its denominator. In reports, in the weekly outbound report, in dashboards, in the sentence you say on a call. "27% on 186 sends" and "27%" are two different claims, and only one of them can be checked.

A fourth rule earned its place in this refresh. Recompute the same number on a new window with the same definition, and print both. Our accept rate moved from 21.5% on 734 invitations to 24.6% on 782, and our reply rate moved from 20.3% on 187 people to 18.4% on 380. Two rates drifting in opposite directions on a doubled base is the normal shape of a pipeline that is growing, and it stays invisible to anyone who quotes one number from one good month.

The part nobody writes down

The work behind these numbers is mostly building lists rather than writing copy. Sourcing people who deserve to be written to, checking a fact per person, watching a sender's health, answering replies within a day, closing the loop between what went out and what came back. That loop is the product we sell, and it is why the numbers above hold on a boring Tuesday as well as in a case study.

If you want to see the same loop running on your own pipeline, the MCP access puts it inside your own Claude: ask who is queued for next week, read what went out in your name, cross out anyone before a message goes.

Questions buyers ask