What is the LinkedIn lead generation funnel?

It is the sequence from a defined audience through to a booked meeting, but there are actually three of them and they are not interchangeable. The outbound funnel starts with connection requests sent. The inbound content funnel starts with post impressions. The paid funnel starts with ad impressions and Lead Gen Form opens. Each divides its conversion rates by a different denominator, which is the main reason published LinkedIn benchmarks disagree so violently with each other.

How many connection requests does it take to book one meeting?

Working from published data as far as it goes: 1,000 requests produce roughly 285 accepted connections at Expandi's 28.5% rate, and roughly 30 replies at their 10.4% reply rate per message sent. Beyond that point no vendor publishes anything. If you assume one in three replies is genuinely interested and half of those book, 1,000 requests produce about 5 meetings. Those last two steps are planning assumptions, not benchmarks, and you should replace them with your own numbers as soon as you have them.

Why do published LinkedIn funnel benchmarks disagree so much?

Almost always because of denominators. A reply rate calculated against accepted connections looks roughly three times better than the same campaign measured against messages sent, and most published figures never state which one they used. Belkins reports 7.2% and Expandi reports 10.4% for a heavily overlapping population over a similar period, and both define it as replies divided by messages sent. Any figure quoted without its denominator is not usable.

How long before a LinkedIn funnel produces meetings?

Expect first conversations in two to three weeks and a steadier flow from around week five. The first fortnight goes on targeting, profile work and message testing. Campaigns that launch faster than that are skipping the part that determines whether the rest works, and the funnel usually stalls at the acceptance stage as a result.

Can I just send more connection requests to book more meetings?

Only up to a point, and that point arrives quickly. LinkedIn does not publish numeric limits, but activity is policed and aggressive sending is what gets accounts restricted. In practice the top of the outbound funnel has a hard ceiling per account, which means scaling means more sending accounts or better conversion further down, not simply more volume from one profile.

Stage by stage

The LinkedIn lead generation funnel,with the math

Every guide draws one funnel. There are three, they use different denominators, and blending them is exactly why the published conversion rates you have been benchmarking against disagree with each other.

Have us run the funnel

What it costs

Short answer

1,000 connection requests produce roughly 285 accepted and 30 replies using published rates. Below that point no vendor publishes anything, so the meeting count depends on assumptions you should replace with your own numbers.

01Three, not one

Three funnels that share a platform and nothing else

These are three different machines. When a guide averages them into one diagram, every rate in it describes a campaign nobody actually ran.

Outbound connection requests

Starts at

Connection requests sent

Denominator

Everything divides by requests sent

The only funnel you fully control the top of, and the only one with a hard ceiling on volume.

Inbound content

Starts at

Post impressions

Denominator

Everything divides by impressions

Slower to start and impossible to forecast, but the replies arrive warm and cost nothing per unit.

Paid Lead Gen Forms

Starts at

Ad impressions

Denominator

Everything divides by impressions, then by form opens

Predictable volume, prefilled forms, and the lowest intent per lead of the three.

02The arithmetic

What 1,000 connection requests actually produce

Drawn exactly to scale, because the collapse is the whole story and most funnel diagrams hide it by making every stage the same size.

Note where the sourced data stops. The top two conversions come from datasets of millions of requests. The bottom two come from assumptions, and we have labelled them as such because every other page on this topic presents the same guesses as benchmarks.

03Every rate, with its source

Stage by stage, and where the data runs out

LinkedIn outbound funnel rates with sources and sample sizes
Stage	Published rate	Source, sample and denominator
Requests sent	Your input	Capped by LinkedIn activity limits, not by budget
Accepted	28.5%	Expandi, 13,218,869 requests, May 2025 to Apr 2026
Accepted (second reading)	26%	Belkins, 2025 dataset
Replied	10.4%	Expandi, 6,730,447 messages. Note: divided by messages sent, not by accepted
Replied (second reading)	7.2%	Belkins, 15.1M contacts, replies divided by messages sent
Positive reply	Not published	No vendor publishes a positive-reply rate. Anyone quoting one is estimating
Meeting booked	Not published	No vendor publishes a reply-to-meeting rate either. Instrument your own

Figures checked August 2026. The two greyed rows are not omissions. No vendor publishes those rates, and we would rather leave a gap than fill it with a number we invented. More on how these datasets relate in our stack breakdown.

04The ceiling nobody models

You cannot fix a funnel by sending more

Run the arithmetic backwards. If 1,000 requests produce around five meetings, then ten meetings a month needs roughly two thousand requests a month from a single profile.

LinkedIn does not publish numeric limits and we will not pretend to know them, but activity rate is policed and aggressive sending is what gets accounts restricted. In practice one profile cannot sustain that volume safely. Scaling means more sending accounts, or better conversion below the acceptance stage. It does not mean turning the dial up.

This is why the honest answer to “can we double the meetings” is usually about targeting rather than volume. We cover what is safe to automate and what is not on our LinkedIn automation page.

05Diagnosis

Five ways this breaks, and what each looks like in the numbers

Campaigns rarely fail everywhere at once. They fail at one stage, and the stage tells you the fix.

Low acceptance

You are in the wrong audience, not using the wrong words. Acceptance ranges 17.5% to 40.1% by industry, while the sender's seniority moves it about 3 points.

Good acceptance, no replies

The first message is pitching. Acceptance says your targeting is fine, so the problem sits entirely in what you sent afterwards.

Replies that go nowhere

Usually speed. Interested replies cool within hours, and most teams have nobody whose actual job is answering them the same day.

Meetings that do not show

Qualification happened too late or not at all. A meeting booked from a polite reply is not a meeting with a buyer.

It worked, then stopped

You exhausted the segment. A tightly defined list is a finite resource, and the fix is a new segment rather than more volume into the old one.

The full six-stage picture, and how we run it as a service, is on the LinkedIn lead generation page. What you would pay for it is on the B2B lead generation cost breakdown, and what you are buying from a provider is set out on the LinkedIn outreach agency page.

06FAQ

The LinkedIn funnel, answered

It is the sequence from a defined audience through to a booked meeting, but there are actually three of them and they are not interchangeable. The outbound funnel starts with connection requests sent. The inbound content funnel starts with post impressions. The paid funnel starts with ad impressions and Lead Gen Form opens. Each divides its conversion rates by a different denominator, which is the main reason published LinkedIn benchmarks disagree so violently with each other.
Working from published data as far as it goes: 1,000 requests produce roughly 285 accepted connections at Expandi's 28.5% rate, and roughly 30 replies at their 10.4% reply rate per message sent. Beyond that point no vendor publishes anything. If you assume one in three replies is genuinely interested and half of those book, 1,000 requests produce about 5 meetings. Those last two steps are planning assumptions, not benchmarks, and you should replace them with your own numbers as soon as you have them.
Almost always because of denominators. A reply rate calculated against accepted connections looks roughly three times better than the same campaign measured against messages sent, and most published figures never state which one they used. Belkins reports 7.2% and Expandi reports 10.4% for a heavily overlapping population over a similar period, and both define it as replies divided by messages sent. Any figure quoted without its denominator is not usable.

Find out where

yours leaks.

A free audit of your funnel stage by stage, using your numbers rather than published averages. You keep the analysis either way.

Book a free pipeline audit

Thirty minutes. You will leave knowing which stage is actually costing you meetings.