---
title: "LinkedIn Ads Cost for B2B in 2026: CPL, CPC, CPM and the Full Distribution"
url: https://metadata.io/linkedin-ads-cost-b2b
description: "What LinkedIn ads actually cost B2B advertisers: $202 per lead weighted, $376 at the median, and the full spread from p10 to p90. 138 advertisers, $57.6M of 2025 spend."
source: metadata.io
---

Channel cost reference · 2025 data

# LinkedIn ads cost for B2B: CPL, CPC, CPM and the full distribution

Every number below comes from 138 B2B advertisers running LinkedIn in 2025, inside a dataset of $57.6M of measured spend with closed-won revenue traced back to the ads. No email gate. The [method is published](https://metadata.io/benchmark/methodology), the [dataset is available](https://metadata.io/benchmark/dataset), and every figure carries the number of advertisers behind it.

## How much do LinkedIn ads cost for B2B?

A B2B lead from LinkedIn costs **$202** on a spend-weighted basis, at $9.39 per click and $63.19 per thousand impressions. Click-through rate is 0.67% and 5.9% of clicks convert. Across 138 advertisers, 2025.

| Metric | Value |
|---|---|
| Cost per lead | $202 |
| Cost per click | $9.39 |
| Cost per thousand impressions | $63.19 |
| Click-through rate | 0.67% |
| Click to lead conversion | 5.9% |

## What is a good LinkedIn cost per lead?

The weighted $202 sits at the 26th percentile. The median advertiser pays $376, nearly double. If you are under $196 you are in the top quarter. Above $658 you are in the worst quarter.

This is the number every other benchmark hides. A single weighted average makes most advertisers think they are failing against it, because a handful of very efficient large spenders drag it down. The spread is the useful part:

| Percentile | Cost per lead | What it means |
|---|---|---|
| p10 | $109 | Top 10% of advertisers |
| p25 | $196 | Top quarter |
| Median | $376 | The typical advertiser |
| p75 | $658 | Bottom quarter |
| p90 | $1,340 | Bottom 10% |
| Spend-weighted | $202 | 26th percentile, not typical |

## Which LinkedIn ad format produces the cheapest leads?

Document ads, at $142 per lead against $200 for image and $265 for video. Document ads also convert at 11.9%, roughly double the channel average. Conversation ads are the most expensive at $362.

| Format | Cost per lead |
|---|---|
| Document | $142 |
| Image | $200 |
| Video | $265 |
| Conversation | $362 |

## Do LinkedIn lead-gen forms beat landing pages?

Yes, and by more than most teams expect. A LinkedIn lead-gen form produces a lead for $193. Sending the same click to a landing page costs $346, 79% more. Measured across 97 form advertisers and 65 landing-page advertisers.

## Is LinkedIn retargeting cheaper than prospecting?

No. On LinkedIn, retargeting costs $234 per lead against $194 for prospecting. The warm audience is the dearer one, which is the opposite of the pattern on Meta. Cold LinkedIn audiences run $194 at 6.6% conversion across 94 advertisers.

## What does a LinkedIn customer actually cost?

Cost per lead is not cost per customer. Across 114 advertisers with closed-won revenue traced to their LinkedIn ads, 21.4% of leads closed and the acquisition cost was $63,312. Triggered return on ad spend was 0.57x.

For companies under 200 employees the picture is different again: $35,288 per customer at a 19.5% close rate, across 27 advertisers. Company size moves this number more than any creative decision does.

## Where does this sit against other channels?

LinkedIn is the most expensive per lead of the social channels and the cheapest per lead against Google Ads. The full cross-channel comparison, with CTR, CPC and CPM for every channel, is on the [B2B advertising benchmarks](https://metadata.io/b2b-advertising-benchmarks) page. The [2026 benchmark report](https://metadata.io/benchmark-report-2026) lets you filter all 111 cuts by company size, industry and campaign goal.

## How were these numbers calculated?

Every figure is a spend-weighted aggregate across advertisers who ran LinkedIn in 2025. Suppression thresholds mean no cut is published on a sample too small to be worth anything; cuts that failed were killed, not caveated. The [full method](https://metadata.io/benchmark/methodology) names the rules and the [dataset](https://metadata.io/benchmark/dataset) is CC BY 4.0.