Scout With Mebloguse cases
ENPTETRU

experiment · · 6 min read

Do 2nd-degree LinkedIn connections accept more requests than 3rd-degree? Only with 5 or more mutual connections: 89%, against about half for everyone else

243 connection requests without a note, each to a stranger whose degree was recorded before the request. Two weeks on, 2nd-degree people had accepted 57% and 3rd-degree people 55%. The label hardly separated anyone; the number of mutual connections did: with 5 or more, 25 of 28 accepted. With only one, they accepted less often than the 3rd degree. Once accepted, none of it changed who replied.

by · Scout With Me

Every name in a LinkedIn search comes with a small label: 2nd or 3rd. The usual advice is to stay with the 2nd, because someone you know knows them, so you arrive a little less of a stranger. It sounds reasonable, and it leaves a third of every list out.

Between September 15 and November 7 I sent 243 connection requests from my own account, without a note, to people I had chosen from their profiles. Before each request the search result was saved, with the degree and the line about mutual connections. So the label could be checked against what people actually did.

What do 2nd and 3rd degree mean on LinkedIn?

They are distances in your network: a 2nd-degree person is connected to one of your connections, and a 3rd-degree person is connected to one of your 2nd-degree connections, as LinkedIn defines them in Your network and degrees of connection (read November 9, 2026). Of my 243 requests, 160 went to 2nd-degree people and 80 to 3rd-degree people.

The label shows next to the name in search results and on the profile. Here it was always read before the request, never after the person accepted, when everyone becomes 1st degree. For a 2nd-degree person, the search result usually also shows how many connections you share; for a 3rd-degree person there are none, by definition.

Do 2nd-degree connections accept more connection requests?

Hardly: of the requests at least two weeks old on November 8, 2026, 2nd-degree people accepted 67 of 117 (57%) and 3rd-degree people 35 of 64 (55%). A gap that size, between groups this size, is well within what chance produces.

I expected the label to separate who accepts from who ignores, and it did not. What separated them was on the line below it in the search result.

How many mutual connections make a difference?

Five or more: of the 2nd-degree people who shared 5 or more connections with me, 25 of 28 accepted the request (89%), against about half in every other group, counting requests at least two weeks old on November 8, 2026.

Mutual connections in the search result Requests Accepted
None (3rd degree) 52 27 (52%)
One (2nd degree) 35 15 (43%)
2 to 4 (2nd degree) 37 19 (51%)
5 or more (2nd degree) 28 25 (89%)

One mutual connection did no better than none: people with a single one accepted slightly less often than the 3rd degree, within chance. Two to four stayed at the same level. The jump comes all at once from five, and that one is far from chance: a result like it would turn up less than once in a thousand times.

This is where my own network comes in. I am Brazilian, and my network is much larger in Brazil than outside it. Half of the 2nd-degree people I invited in Brazil shared 5 or more connections with me; outside Brazil, 10 of 95 did. The same label meant different things: in Brazil, many people in common; outside, one or two. That is why, split by country, the 2nd degree seemed to pay off in Brazil, 76% against 55% for the 3rd degree, and not at all elsewhere. Within the group of 5 or more, both sides accepted almost alike: 20 of 22 in Brazil and 5 of 6 outside.

Do 2nd-degree connections reply more after accepting?

No: of the people who accepted and received a first message by November 7, 2026, 52 of 88 in the 2nd degree replied (59%) and 28 of 40 in the 3rd degree (70%), counted on November 8. Mutual connections did not change that either.

Mutual connections in the search result Accepted and got a message Replied
None (3rd degree) 30 21 (70%)
One (2nd degree) 20 9 (45%)
2 to 4 (2nd degree) 30 18 (60%)
5 or more (2nd degree) 26 17 (65%)

People who share many connections with you accept almost every time, and then reply at the same rate as people who share none. The network gets the request accepted; the conversation depends on what you write next. How fast they accepted did not change it either: Do people who accept a LinkedIn request fast reply more?

What to do with this

  1. Keep 3rd-degree people on the list. A third of my requests went to them; they accepted at the same rate as the 2nd degree and, once connected, replied a little more often.
  2. If you prioritize, prioritize by the number of mutual connections, not by degree. With 5 or more, almost everyone accepted; with one or two, it was like sending to the 3rd degree.
  3. Expect more acceptances where your network is large. That is where the 2nd degree tends to come with many people in common.
  4. Do not expect the network to make people reply. After the acceptance, the people who replied were the ones who found a reason in the first message: What makes a first LinkedIn message get a reply?

Scout With Me reads each profile before the request and writes down why the person is worth a conversation; choosing who gets a request stays yours.

How these numbers were measured

  • Source: Scout With Me running on my own LinkedIn account, prospecting for real, for the products I sell. No data from other users of Scout With Me; the app keeps its data on each user's computer and sends nothing to us.
  • What was counted: request, a connection request without a note to someone who was not a connection; degree, read from the search result saved before the request, or, for 24 requests with no saved result, from the profile read before the request; mutual connections, the number the search result line showed before the request; accepted, the request was accepted at any point before the count; replied, the person wrote at least once after accepting and getting the first message.
  • When: requests sent from September 15 to November 7, 2026, counted on November 8. Acceptance uses only requests at least two weeks old, sent by October 24, because younger ones may still be pending; replies use everyone who had accepted and received a first message by November 7.
  • Size and chance: 243 requests: 160 to the 2nd degree, 80 to the 3rd, 3 with no degree recorded, left out. The mutual-connection tables use only requests with a saved search result; 11 mature 2nd-degree requests whose result showed no mutual-connection line are also left out of them (4 accepted). Two-sided Fisher's exact test: 2nd against 3rd degree, acceptance p = 0.76 and replies p = 0.32; 5 or more in common against the rest of the 2nd degree, acceptance p < 0.001; against the 3rd degree, p = 0.001; one in common against the 3rd degree, p = 0.51. Outside Brazil, the group of 5 or more has only 6 requests. Rates are rounded to whole percentages.
  • Left out: five people who were already connections and got the first message without a request. Nobody is named, no message is quoted.
  • The rules that apply to every post are in How the numbers on this blog are measured.

Changelog

  • 2026-11-09: first version. 243 requests sent from September 15 to November 7, 2026, counted on November 8.

read this post as Markdown