TL;DR: Accounts with 5,000 or more LinkedIn connections saw 34.9% acceptance against 23.2% for the 500-999 band, but the gap is not causal because successful sending is what grows the network.
Our analysis of 98,500 matured LinkedIn connection requests found accounts now holding 5,000 or more connections had 34.9% of their cold requests accepted against 23.2% for accounts in the 500 to 999 band, a gap that cannot be read as causal because successful sending is itself what grows a network. Our analysis
Key Takeaways
- Accounts now holding 5,000 or more connections accepted 34.9% of cold requests, against 23.2% for the 500-999 band.
- The under-500 band beat the 500-999 band by two points, which is not what the "thin profile" folklore predicts.
- Connection counts here are measured at the pull date, not the send date, so successful accounts have already been moved into higher bands by their own history.
- The under-500 band still accepted at 25.29%, within two points of the cohort's overall matured rate.
- The joined subset accepts at 29.83% against the cohort's 26.97%, meaning surviving accounts skew successful, a selection effect this study inherits.
What did we measure?
"Do I need to build my profile before starting outreach" is a question every new sender asks, and the standard answer, five hundred connections minimum, is folklore with no denominator behind it.
This study measures the association that folklore points at: acceptance rate by the sending account's connection count, across every matured request that joins to a sending account with a recorded count. It then spends as much space on why the association is not causal as on the association itself, because this is the study we most often see done dishonestly elsewhere.
How was it measured?
This study checks whether a sending account's current connection count relates to how often its requests get accepted.
Dataset window. Connection requests sent between 8 January 2025 and 27 July 2026, pulled 27 August 2026, matured at least 30 days.
Coverage. 98,500 requests (50.08% of the matured cohort) join to a sending account with a current connection count; the rest were sent by accounts since removed or without a recorded count. The joined subset accepts at 29.83% against the cohort's 26.97%, meaning surviving accounts skew successful, which is itself a selection effect this study inherits. Our analysis
The central limitation, up front. Connection counts are measured at the pull date, not at the send date. An account's requests from a year ago are attributed to the band it occupies today. Since accepted requests become connections, an account that sent successfully for a year has been moved up bands by its own success. This is why the study carries ANALYSIS provenance and why no sentence below claims a big network causes acceptance.
Band audit counts. Per-band sending-account counts are published so the bands can be audited; they are analysis subset sizes, not a population claim.
What did we find?
The table below groups matured requests by the sending account's connection count at the time of the pull, not at send time.
| Sender's connections (at pull) | Requests | Sending accounts | Acceptance | Largest workspace's share |
|---|---|---|---|---|
| Under 500 | 3,159 | 10 | 25.29% | 35.74% |
| 500-999 | 3,762 | 6 | 23.21% | 34.90% |
| 1,000-1,999 | 34,817 | 24 | 29.07% | 31.41% |
| 2,000-4,999 | 32,378 | 17 | 28.04% | 19.24% |
| 5,000+ | 24,384 | 10 | 34.89% | 30.91% |
The spread is 11.68 points and it is not monotonic: the under-500 band beats the 500-999 band by two points. That non-monotonicity matters, because the folklore's mechanism (a thin profile scares people off) predicts the smallest accounts should do worst, and they do not. Our analysis
What does this mean for new LinkedIn senders?
Three explanations fit this table, and the data cannot pick between them.
Reverse causality. Accounts at 5,000+ connections today include the accounts that have been sending successfully the longest. Their history built the band they now sit in. Under this reading the table mostly measures time-in-service and skill, and the network is the exhaust, not the engine.
Real social proof. A visible 500+ or 5,000+ connection count may genuinely reassure recipients, and larger networks produce more mutual connections, which LinkedIn displays prominently on a request. Under this reading some of the gap is causal.
Shared-audience effects. Big-network accounts in this dataset skew toward senders whose audiences already know them. Their acceptance advantage would then be a targeting property, not a profile property.
The honest summary is that big-network senders do better and nobody should promise you that growing your network will make you do better. What the table does refute is the strong version of the folklore: the claim that outreach from a small account is pointless. The under-500 band, ten accounts and 3,159 requests, accepted at 25.29%, within two points of this cohort's overall matured rate of 26.97% (the earlier benchmark hub reports 27.11% on its own frozen cohort). Our analysis
Senders who would rather target well than wait to grow a profile can see how Reachium's plans support targeting from day one.
How do you use this in practice?
A founder with a fresh 300-connection profile deciding whether to wait and "build presence" first has a measured answer for the waiting's cost: none visible here. The under-500 band's 25.29% is not the table's best rate, but it is a working rate, and acceptance compounds into meetings within days, not months. Our analysis
An agency promising clients higher acceptance after "profile warming" to 1,000+ connections is selling the 500-999 to 1,000-1,999 jump, 5.9 points in this table. That jump is real in the cross-section and unproven as an intervention; this study is the citation for the difference between those two statements. Our analysis
Tools that market network-size and reach as the core pitch are selling the same variable this study says is confounded with tenure. Reachium's Sales Navigator comparison covers where the two tools differ.
What are the limits of this study?
This study's biggest limitation is timing: connection counts are captured after the fact rather than at the moment each request was sent.
Counts are current, not historical. The single largest threat to validity, described in Methodology. A longitudinal cut (each request against the sender's count on its send date) is the right design and this dataset does not store the history to run it.
Survivor selection. Only accounts still connected at pull time appear, and they skew successful (29.83% joined subset acceptance against 26.97% platform). Our analysis
Small low bands. The two bands under 1,000 connections hold 16 accounts and under 7,000 requests combined; their rates carry wide uncertainty, and monthly variance in this data is large relative to a two-point difference. Platform data
Concentration. Per-band largest-workspace shares are published above; the under-1,000 bands lean hardest on single workspaces.
Connections, not followers. Follower counts are not analyzed; for accounts running content alongside outreach the two diverge, a dynamic the follower-growth study measures from the other side. Teams weighing whether to invest in profile growth or targeting can talk to the agency team about which lever fits their stage.
FAQ
Do you need a big LinkedIn network to run outreach?
No. Accounts with under 500 connections accepted 25.29% of cold requests, within two points of the overall matured rate, so a small network is not a barrier to starting.
Why did accounts with 5,000+ connections have a much higher acceptance rate?
Partly because successful sending is what grows a network in the first place. Accounts sending well for a long time end up in the high bands, so the gap likely reflects tenure and skill as much as network size itself.
Should a new LinkedIn account wait to build up connections before starting outreach?
This data does not support waiting. The under-500 band performed close to the cohort average, and the biggest acceptance drivers measured elsewhere are targeting factors, not network size. Teams can start targeting well from day one on Reachium's pricing page.
Does a bigger LinkedIn network cause higher acceptance rates?
This study cannot say. Reverse causality, real social proof, and shared-audience effects all fit the data, and nothing here isolates network size as the cause.