LinkedIn prospecting in 2026: what we learned from 4,086 invitations and 1,053 messages

Prospects who engaged with a post accept almost twice as often

A prospect who liked a LinkedIn post accepts the invitation 60.5% of the time (3 accounts), one who commented 49.6%, against 28.1% for a profile found through search. Liked and commented together: 54.6%, and that interval does not overlap the search one. Authors of recent posts on your keywords, which Makelead spots for you, sit in between at 40.0%.

Likers alone come from only 3 accounts, and a gap between accounts could pass for a source effect. So we also compared both sources inside the same account. 4 accounts sent at least 30 invitations from each. In the 3 customer accounts, people who liked or commented on a post accepted 2.6, 2.3 and 2.4 times more often than search results. In the founder's account the gap is smaller, 1.3 times, because its search results already accept at 49.2%.

Replies lean the same way: 39.8% of messaged prospects who came from a post replied, against 22.0% from search. Here the intervals overlap, and in customer accounts alone the gap narrows, so read it as a trend rather than a proof.

To build a list from the people who react to posts in your niche, see how to turn post engagement into leads.

One follow-up pays off, the second much less

60 of the 290 replies (20.7%) came after a follow-up. The first follow-up got 10.5% of still-silent prospects to reply (53 of 506), the second 3.9% (7 of 178). In Makelead, the sequence stops at the first reply.

Tools that handle follow-ups: LinkedIn sequence and follow-up tools.

It all happens within hours

Half of accepted invitations were accepted in under 9 hours (median 8.9 hours), and 75% within 2 days. On the message side, half of replies arrived less than 5.5 hours after the first message (290 replies measured), 75% within just over 3 days. An invitation still pending after a week has little chance left.

The day you send barely matters

By day of sending, acceptance ranges from 28.0% on Tuesday to 37.4% on Wednesday, and every interval overlaps the others. Weekend invitations do at least as well as weekday ones. As for the hour, 79% of invitations left between 8 am and noon (Paris time), so hours cannot be fairly compared with this sample.

A longer first message does not get more replies

The median first message is 212 characters long. The shortest quarter (under 173 characters) got 32.5% replies, the longest quarter (258 and over) 23.1%. In between, rates do not follow length (25.5% then 28.7%) and the intervals overlap. So we cannot say short wins, only that adding length does not buy replies.

To write it: the AI LinkedIn message writer.

What to take away

• Invite first the people who reacted to a post in your niche. That is where acceptance nearly doubles, inside the same account too.
• Plan one follow-up. About one reply in five comes after a follow-up, and the second gets 2.7 times fewer still-silent prospects to reply than the first.
• Do not hunt for the perfect day: the target moves the numbers, the calendar does not.
• Keep the first message short and about the person. The median is 212 characters.

Makelead runs this for you: it finds the people who engage with posts in your niche, sends invitations within weekly caps, sends your messages and follow-ups (written by AI on the Growth and Business plans), and stops the follow-ups as soon as the prospect replies. See LinkedIn prospecting with Makelead or how it compares to Waalaxy.

Method and limits

• Data: every LinkedIn invitation sent through Makelead from July 25 to October 25, 2026 (93 days, Paris time) by 25 accounts, the founder's included, read from our database on October 26, 2026. Acceptances and replies are counted up to that day. Without internal accounts (founder, Makelead addresses, tests), the 23 customer accounts give 30.8% acceptance (1,121 of 3,638), 24.9% replies, and 48.1% for post engagers against 26.7% for search.
• Acceptance rate: invitations marked accepted divided by invitations sent. Reply rate: prospects who wrote back after the first campaign message (interest, refusal or question alike), divided by prospects who received that message.
• Follow-ups: the date of the first reply is compared with the send dates of the first message and of each follow-up.
• Intervals: 95%, by bootstrap over accounts (2,000 draws). Prospects of the same account are not independent, so the intervals are wide on purpose.
• Threshold: no figure published below 30 cases. Below 5 accounts, the number of accounts is written next to the figure.
• Privacy: aggregates only. No name, no profile, no message text (only its length), no identifiable account.
• Limits: 25 accounts, mostly French-speaking freelancers, consultants and small agencies. The largest account weighs 10.6% of invitations, the 3 largest 30.5%. The median acceptance rate per account is 39.6% (half of the 20 accounts with 20+ invitations sit between 25.4% and 47.4%).
• This is observational, not an experiment: a gap is a correlation, not a proven cause. Recent invitations have had less time: on invitations older than 14 days, acceptance is 35.9%.
• Not published: connection degree (our database overwrites it once the invitation is accepted, so any rate by degree would be biased) and invitation notes (only 240 invitations, on 4 accounts, had one).

Frequently asked questions