How Many LinkedIn Posts a Day Is Too Many?
Last updated · Research by the B2Ne team
In our data, more than one. In October 2026 we tracked 42 posts on the LinkedIn profile of Mehtab Singh, B2Ne's founder, while changing how often we posted and which AI wrote the copy.
The results at a glance
Days with 1 post beat days with 3 on every measure below.
- 1 a day
- 3 a day
Impressions for a typical post
Share of impressions from outside your network
Posts that got any reaction, comment or repost
Also: across all 42 posts, nobody saved one, sent one or followed after one.
42 posts on one founder's profile, 1–31 October 2026. Typical = the middle post (the median), explained below.
See the numbers
| 1 a day | 3 a day | |
|---|---|---|
| Posts | 8 | 24 |
| Median impressions per post | 85 | 60 |
| Out-of-network share for the group | 32.9% | 21.7% |
| Posts with at least one social engagement | 7 of 8 | 10 of 24 |
| Saves, sends on LinkedIn and followers gained (all 42 posts) | 0 | 0 |
How did we run the study?
We changed how often we posted on one profile in one month, and compared the posts.
- Claude campaign post
- Gemini campaign post
- Gemini article link post
- Human-written
Claude's posts mostly went out three a day in early October. Gemini's posts mostly went out one a day in late October.
Source: B2Ne October 2026 LinkedIn study
Claude wrote the copy for the three-a-day campaign and Gemini for the one-a-day campaign. Mehtab also wrote 3 posts himself, so we could compare AI-written posts with human-written ones. Gemini's 3 article link posts count in the AI-written total but not in the Claude vs Gemini comparison, because a link post has a different job.
See the setup in a table
| Setup A | Setup B | Human-written | |
|---|---|---|---|
| Who wrote the copy | Claude | Gemini | Mehtab, writing himself |
| How often | Mostly 3 campaign posts a day (24 of 28 posts) | Mostly 1 campaign post a day (7 of 8 posts) | 1 post a day |
| Dates | 6–16 October, plus 1–2 October | 19–28 October, plus 1 October | 19–23 October |
| Posts measured | 28 | 8 | 3 |
| Formats | Images, videos, carousels | Images, videos, a carousel | Text posts and a reshare with his comment |
- We read each post's numbers from its LinkedIn analytics page, 7 days after posting where we could.
- People set the questions and checked every finding. An AI assistant copied the numbers from LinkedIn into our tracking sheet and worked out the medians.
How do you read the numbers?
We use LinkedIn's own labels, and the median for each group of posts.
- Impressions: how many times a post appeared on someone's screen. One person scrolling past twice counts twice.
- Members reached: how many different people saw it.
- Out-of-network: the share of impressions from people who don't follow Mehtab and aren't connected to him. New buyers come from here. The rest is In-network (followers and connections).
- Social engagements: reactions, comments and reposts added together.
- Median (the middle post): line the posts up from lowest to highest and take the one in the middle. With an even number of posts, it is halfway between the two middle ones, which is why some medians end in .5.
Impressions of the 8 posts on one-a-day days
Source: B2Ne October 2026 LinkedIn study
We use the median because one big post can pull the average up. The 8 one-a-day posts got 61, 67, 75, 84, 86, 104, 149 and 259 impressions: the median is 85, but the average is 111 (885 ÷ 8), even though 6 of the 8 posts got fewer than 111.
Does posting more often hurt each post?
Yes, in our data: the median post got 85 impressions on one-a-day days and 60 on three-a-day days, and a smaller share of those impressions came from outside the network.
Median impressions per post, by campaign posts that day
Source: B2Ne October 2026 LinkedIn study
Share of impressions, In-network vs Out-of-network
- In-network (followers and connections)
- Out-of-network
Source: B2Ne October 2026 LinkedIn study
See the numbers
| Campaign posts that day | Posts | Median impressions | Median members reached | Median out-of-network impressions | Out-of-network share for the group |
|---|---|---|---|---|---|
| 1 | 8 | 85 | 30 | 17.4 | 32.9% |
| 2 | 4 | 72.5 | 26.5 | 13.6 | 19.2% |
| 3 | 24 | 60 | 25.5 | 10.7 | 21.7% |
Source: B2Ne October 2026 LinkedIn study, 36 AI-written campaign posts. Out-of-network impressions = impressions × LinkedIn's Out-of-network share, per post. The share for a group = all its out-of-network impressions ÷ all its impressions. The 2-a-day group has only 4 posts.
Three posts a day brought more impressions in total, because there were more posts. But each post got fewer, and more of them stayed inside the network Mehtab already had.
Larger studies point the same way. AuthoredUp's analysis of 3 million LinkedIn posts found a median of 586 impressions per post for profiles posting 8 or more times a week, against 870 at 4–5 times a week. FlyPost's study of 116,000+ posts from 2,400+ B2B creators found that a post that follows another within a day earns about 10% fewer reactions.
Did anyone act on the posts?
More of the one-a-day posts got a response, but no post got a save, a send or a new follower.
Posts with at least one reaction, comment or repost
1 a day: 7 of 8 posts (88%)
3 a day: 10 of 24 posts (42%)
Source: B2Ne October 2026 LinkedIn study
See the numbers
| 1 a day | 3 a day | |
|---|---|---|
| Posts with at least one social engagement (reaction, comment or repost) | 7 of 8 (88%) | 10 of 24 (42%) |
| Social engagements per 100 impressions | 0.79 | 0.98 |
| Profile viewers from these posts | 3 | 7 |
| Saves, sends on LinkedIn and followers gained | 0 | 0 |
Source: B2Ne October 2026 LinkedIn study, AI-written campaign posts.
For every 100 impressions, people responded at about the same rate on both kinds of day. The difference is spread: on one-a-day days nearly every post got a response, while on three-a-day days most posts got none.
Who saw the posts?
Senior people at small IT services companies were the largest viewer group, whoever wrote the post.
LinkedIn's Top demographics shows only the largest viewer group for each post in each category. We counted how many posts each group led.
LinkedIn Top demographics: how many posts each group led
Claude, 3 a day
Gemini, 1 a day
Human-written
Source: B2Ne October 2026 LinkedIn study
See the numbers
| Top demographics (LinkedIn's label) | Claude, 3 a day | Gemini, 1 a day | Human-written |
|---|---|---|---|
| Seniority: Senior | 24 of 25 posts (median share 50.5%) | 5 of 5 posts (45%) | 3 of 3 posts (42%) |
| Company size: 2-10 employees | 21 of 21 posts (33%) | 4 of 6 posts (35.5%) | 3 of 3 posts (35%) |
| Company size: 11-50 employees | none | 2 of 6 posts (22.5%) | none |
| Industry: IT Services and IT Consulting | 9 of 9 posts (24%) | 5 of 5 posts (29%) | 3 of 3 posts (24%) |
| Location: Karlstad | 5 of 7 posts | none | 1 of 1 post |
| Location: Greater Stockholm Metropolitan Area | 1 of 7 posts | none | none |
| Location: London Area, United Kingdom | 1 of 7 posts | 2 of 2 posts | none |
| Job title: Founder or Owner | 2 of 3 posts | none shown | none shown |
Source: B2Ne October 2026 LinkedIn study. "N of M posts" counts only the posts where LinkedIn showed that category. The share in brackets is the median share of impressions that group had on the posts it led. Company names in the viewer data are not shown.
What this means for you: the posts reached people close to a B2B founder's buyer, but founders and owners led on only 2 posts, so reaching the decision-maker by name is still the gap. The one-a-day posts were also led by London viewers on 2 posts and by slightly larger companies (11-50 employees) on 2 posts: a hint they travelled further, not yet a result.
Does it matter whether Claude or Gemini writes the copy?
We can't say yet, because the writer and the number of posts a day changed at the same time.
Campaign posts by writer and by campaign posts that day
almost every Claude post went out on a three-a-day day, and almost every Gemini post on a one-a-day day, so the two effects are tangled together.
Source: B2Ne October 2026 LinkedIn study
See the numbers
| Claude, 3 a day | Gemini, 1 a day | |
|---|---|---|
| Median impressions | 61.5 | 79.5 |
| Median members reached | 26 | 30 |
| Median out-of-network impressions | 11.7 | 16.6 |
| Out-of-network share for the group | 21.6% | 32.8% |
| Posts with any social engagement | 12 of 28 | 7 of 8 |
| Social engagements per 100 impressions | 0.92 | 0.94 |
| Profile viewers from these posts, per 1,000 impressions | 3.8 | 3.5 |
Source: B2Ne October 2026 LinkedIn study, 28 Claude and 8 Gemini campaign posts.
Gemini's posts reached more people, but in the same pattern as the posting-frequency results, so posting frequency is the likelier explanation. For every 100 impressions, both got responses at about the same rate.
How did human-written posts compare with AI-written posts?
Mehtab's own posts got more impressions, but mostly from people who already follow him or are connected to him.
- AI-written
- Human-written
Median impressions
Median members reached
Where the impressions came from
- In-network (followers and connections)
- Out-of-network
Source: B2Ne October 2026 LinkedIn study
See the numbers
| AI-written (39 posts) | Human-written (3 posts) | |
|---|---|---|
| Median impressions | 66 | 94 |
| Median members reached | 26 | 40 |
| Out-of-network share for the group | 23.7% | 9.3% |
| Social engagements per 100 impressions | 0.83 | 0.31 |
Source: B2Ne October 2026 LinkedIn study. AI-written includes the 3 article link posts. In the week of the human-written posts (19–23 October), the median of the 5 AI-written posts was 75 impressions, with 29.2% from outside the network.
With only 3 human-written posts, treat this as a first look.
What should you do with this?
- Post once a day before you post more. Each extra post in a day went with fewer impressions and fewer out-of-network impressions per post.
- Watch your Out-of-network share. It moved more than any other number, and it's on each post's analytics page.
- Make posts people want to keep. Zero saves and sends across 42 posts means people saw the posts but didn't keep or pass them on. A step the reader can check, or a number they can test, is a place to start.
- Choose the AI writer later. Settle how often you post first, then compare writers fairly.
Test it on your own profile in 24 hours: open the analytics of your last 10 posts (View analytics under each post), and write down each post's Impressions and Out-of-network share. Mark the posts from days you posted more than once, then compare the middle post of that group with the rest. If your busy days come out lower, try one post a day for two weeks and compare again.
What can't this study tell you?
It shows what happened on one profile in one month, not what caused it.
- The AI writer, the number of posts a day, and the campaigns, topics and visuals all changed together.
- 16 of the 28 Claude posts were read later than 7 days after posting, so their numbers include extra days. This favours Claude.
- LinkedIn's Top demographics shows only the largest viewer group per post, and "campaign posts that day" leaves out other posts and reposts on the profile.
- There are only 3 human-written posts, and only one profile.
Edition 2, planned for December 2026, gives Claude and Gemini the same brief at one post a day, about 6 posts each, read at day 7, with more human-written posts.
Frequently asked questions
Hand this to whoever runs your company's LinkedIn. If you want a second pair of eyes on your own numbers, we're happy to look at them with you.
Book a 30-Minute Discovery CallRelated reading
- How to Use LinkedIn for B2B Lead Generation
- Why Most B2B Content Gets Ignored, and How to Fix It
- B2B Website Analytics: Six Questions to Answer
- Services: Social Media · Analytics & Insights
How to cite this study
B2Ne (Business To Nexus). How Many LinkedIn Posts a Day Is Too Many?, Edition 1. 13 November 2026. https://www.business2nexus.com/nexus-academy/research/how-many-linkedin-posts-a-day-2026
Method notes
- Posts dated 1–31 October 2026 on Mehtab Singh's LinkedIn profile (2,089 followers on 19 October 2026), read from each post's LinkedIn analytics page.
- 42 posts: 28 with Claude copy, 8 with Gemini copy, 3 Gemini article link posts, 3 human-written.
- Left out: reposts of our own posts, and posts from 29–30 October that hadn't been read yet.