The fastest way to lose a co-author’s trust is to post something they wouldn’t have written themselves. A finding described as bigger than the data supports, a caveat dropped because it slowed the sentence down, a collaborator’s contribution left out of the framing. None of it takes malice, only a Sunday afternoon and a copy of the abstract.
The paper and the post are different documents, written for different readers, and the post takes more care, not less.
Read the methods and the limitations, not the abstract alone
The abstract sells the paper to reviewers who already understand the field. The post has to explain the same result to someone who doesn’t, without the shorthand that makes the abstract efficient. That means going back to the methods section for what the study measured, and the limitations section for what it didn’t show.
A result presented without its limitations reads differently to a reviewer than the paper itself does. If the study was in mice, say mice. If the effect was measured under conditions that don’t hold outside the lab, say that. The people most likely to catch the difference are the people who’ll be reviewing your next grant.
A worked example
Say a lab found that a compound reduced inflammation markers in a mouse model of a chronic condition. An overclaimed post: “Our new treatment reduces inflammation and could change how we treat chronic disease.” Two problems: “treatment” implies something tested in people, and “could change how we treat” promises an outcome the study never measured.
A version that holds up: “In a mouse model, this compound cut a key inflammation marker by a meaningful margin. It’s early, mouse to human is a long road, and the next step is testing whether the effect holds in a different model.” Same result, same enthusiasm about the finding, and every sentence would survive a careful co-author’s read.
Write it, then send it to your co-authors before it goes out
Not after. A co-author who sees a post about their work for the first time in their own feed is going to remember it, and not warmly. Send the draft, ask if the framing is right, and give real time for a reply rather than a same-day deadline. This one habit prevents most of the damage that comes from posting about joint work.
This matters even more for a paper with a trainee as first author. Their name is the one attached to the finding for the rest of their career, and a post that overstates what they found follows them into every future job talk and grant review where someone remembers the claim and checks it against the paper.
Write for the platform, not the abstract
LinkedIn readers scan and decide in the first two lines whether to keep going. Open with the result and why it matters to the field, not a chronology of how the project started. Save the methodological detail for a paragraph in, once the reader is already invested.
A structure that works for most results:
- The finding, stated plainly, in the first sentence
- One sentence on why it matters to people outside the immediate subfield
- A caveat or limitation, stated as directly as the finding
- A link to the paper, not a summary of the entire discussion section
Common ways a post overclaims without meaning to
A few patterns show up again and again, none of them intentional. Describing a correlation as a cause, because “linked to” got shortened to “causes” somewhere in editing. Dropping the sample size, so a result from twelve animals reads like a settled finding. Using “proves” or “shows conclusively” for a single study, when the field would call it one data point among several needed. None of these require bad intent. They require a second read before the post goes out, checking every strong verb against what the data supports.
What to leave out
Skip the exclamation points and the “thrilled to share.” A result that matters doesn’t need enthusiasm attached to it to land. State it, and let the work carry the weight.
Common questions
What's the biggest mistake labs make when posting about a new paper?
Dropping the limitations to make the sentence read cleaner. A finding described as bigger than the data supports is the fastest way to lose the trust of reviewers who read the actual paper.
Should co-authors see the post before it goes out?
Yes, every time, with real time to respond rather than a same-day deadline. A co-author who first sees a post about their own work in their feed remembers it, and not well.
How long should a LinkedIn post about a paper be?
Long enough to state the finding, explain why it matters, and name a limitation, usually four to six sentences. Save the methodological detail for people who click through to the paper.
Is it fine to post about a preprint before peer review?
Yes, but say so directly. Calling a preprint result a paper, or leaving out that it hasn't been reviewed, is the kind of gap a careful reader in your field will notice.