LinkedIn and X reward opposite things, and most founders learn that the expensive way by writing one post and pushing it to both. We measured 1,418 LinkedIn posts and 2,055 X posts across the founder accounts Megaphone runs, ranking every post against its own author's typical post so accounts of different sizes compare fairly. All of it measured in August 2026. The short version: LinkedIn has a higher floor, X has a higher ceiling, and the first line that works on one is often the worst thing you could write on the other. This article covers what the numbers actually say, the weekly split we run, and how to decide if you only have room for one.
The reach shape is completely different
The single most useful thing in the data is not about hooks or length. It is about how reach is distributed.
| LinkedIn (1,418 posts) | X (2,055 posts) | |
|---|---|---|
| Share of impressions from the top 10% of posts | 74% | 92% |
| Share from the top 1% | 29% | 64% |
On X, roughly two thirds of everything you will ever reach comes from one post in a hundred. On LinkedIn the load is spread much wider.
That difference decides how you should think about a bad week. A run of ordinary LinkedIn posts still lands in front of people, because the platform carries the middle of the distribution. A run of ordinary X posts is close to shouting into a room with nobody in it. X pays you in rare, enormous posts, and the price of admission is volume.
Both platforms produce a post at three times the author's usual about one time in six. What differs is what that post is worth when it lands.
The first line does opposite things
This is where copying a post across platforms costs you the most. Both tables rank the first line of a post by how it performed against that author's own typical post.
The reach numbers above come from live-profile records. The first-line tables below come from a deeper history pull across the same accounts, measured in August 2026, which is why the counts differ.
LinkedIn, 2,707 posts:
| First line contains | Median lift |
|---|---|
| Image, document or video attached | 3.21x |
| A dollar figure | 1.69x |
| A story marker ("years ago", "when I") | 1.42x |
| First person (I / My / We) | 1.39x |
| Over 1,200 characters | 1.38x |
| Under 280 characters | 0.63x |
| A question | 0.67x |
X, 2,022 posts:
| First line contains | Median lift |
|---|---|
| A list of three or more items | 1.77x |
| A story marker | 1.63x |
| A dollar figure | 1.48x |
| First person | 1.42x |
| A question | 1.25x |
| Under 280 characters | 0.80x |
| Names a company or person early | 0.75x |
Opening with a question is among the worst things you can do on LinkedIn and a solidly above-average move on X. If you take one operational rule from this article, take that one.
Two more that survived the data. Attaching a visual is the largest single effect we measured anywhere: 3.21x on LinkedIn, and 56 percent of those posts cleared three times their author's usual. And short posts lose on both platforms, which surprises people who think of X as the short-form one. Under 280 characters runs 0.80x on X and 0.63x on LinkedIn.
Leading with someone else's name costs you about a quarter of your reach on X. Lead with your own experience and name the company later in the post.
What we could not confirm
Three claims from the public playbooks did not hold up in our data, and I would rather say so than repeat them.
External links do not cost you 60 percent of your reach on LinkedIn. We measure a 9 percent penalty. Put the link at the end as a preference, not because of a number somebody quoted you.
Comments do not carry fifteen times the weight of likes. The comment-to-reaction ratio is the same on posts that took off and posts that went nowhere.
And posting time barely matters. Monday runs 1.16x on LinkedIn, Sunday 0.91x, and every other day sits within 3 percent of average. Anchor on Monday if you want an easy decision, then stop thinking about timing.
The split we actually run
Five X posts, three LinkedIn posts, one newsletter issue, every week. That weekly plan is the content layer of the system we install for clients, and it is the same one I run on my own accounts, which are at 93,000 on X and 26,000 on LinkedIn.
The shape follows the reach data. X gets more posts because volume is how you buy chances at an outlier. LinkedIn gets fewer and more considered ones, with a visual on every single one, because each post has a real floor and the visual is worth more than anything else you could add.
Two of the five X posts should be long. Over 1,200 characters runs 1.33x there, and 31 percent of those clear three times the author's usual. Only list openers do better, at 38 percent.
Whatever split you choose, the number to protect is the floor. Under two posts a week, no account in our book grew more than 8 percent, with one exception. Above two, most grew 45 percent or more. Volume alone is not enough either, since we have an account posting five times a week that grew 6 percent, so the schedule is necessary and not sufficient.
If you only have room for one
Pick the platform where you can name five real people who would read you: an investor you want in the next round, two operators your candidates follow, the person your customers quote.
For most classic B2B companies that is LinkedIn, because the buying committee is there in work mode. For developer tools, AI infrastructure and anything sold to very online people, it is X. The platform's general merits are irrelevant next to where your specific customers already read.
If LinkedIn is the honest answer, our honest answer on whether posting on LinkedIn works for B2B covers the conditions that decide it. If you are raising, how investors read LinkedIn during fundraising due diligence is the piece to read next.
The verdict
LinkedIn is the floor and X is the ceiling. Run both if you can hold the schedule, because they fail in different weeks and a quiet month on one is usually not a quiet month on both. Write for each platform separately: a visual and a story on LinkedIn, a list or a long thread on X, and never the same first line in both places.
If you want to know which of the two is actually costing you right now, the feed diagnostic is six questions and names the one thing to fix first.