Sharing Your Work on PACR: Posts That Get Noticed

Most research is finished in private and announced in public, and that order is backwards.

We spend two or three years on a project, talk to maybe a dozen people about it while it's still soft and changeable, then publish it and post a link that says "New paper out!" with a celebration emoji. A few colleagues like it. Nobody argues with it. Six months later the most useful conversation you'll ever have about that work happens by accident at a poster session, and it's too late to change anything.

I've written a lot of those link posts. I've also watched, from inside academia and now from the product side of PACR, an AI-native research workspace, what actually gets researchers talking to each other. The difference rarely comes down to how important the work is. It's how the post is built.

This piece is about building better posts on PACR: ones that get read, get answered, and occasionally change what you do next.

Why "new paper out" posts go quiet

A link post asks the reader to do all the work. They have to click, find the abstract, decide whether it matters to them, and then figure out what they'd even say in response. Most people stop at the first step and tap like on the way past, if they bother at all.

The other problem is subtler. A finished paper feels closed. The tone of an announcement says "this is settled," and people are polite about settled things. Comments come when a reader sees an opening: a question they can answer, a choice they'd have made differently, a result that contradicts their own data.

There's a decent body of work on whether promoting papers on social platforms actually moves citations or readership, and if you search PubMed for trials of social media promotion and citations, you'll find the results are more mixed than the enthusiasts suggest. I read that as good news. Broadcasting isn't the lever. Conversation is.

What a grant officer taught me about being remembered

Years ago I sat through a funder's workshop for early-career applicants. The program officer running it, Marcus, had read thousands of applications. Someone asked him what made a proposal memorable, expecting an answer about preliminary data or a strong team.

He said the ones he remembered were the ones he could explain to his own director in the time it took to walk to the elevator. One sentence of what the applicant wanted to find out, one sentence on why nobody had managed it yet. If he couldn't do that, the proposal blurred into the pile no matter how good it was.

I've never forgotten it, partly because my own proposal at the time failed that test badly. And I think the same rule governs posts. If someone scrolling past can't repeat your point to a colleague, they won't engage with it, because they don't yet know what "it" is.

So before you write anything on PACR, write the elevator sentence. Keep it out of the post if you like, but know it.

Formats that actually start conversations

These are the formats I've seen work again and again. None of them is clever. They work because each one leaves a door open for the reader.

The surprise post

Lead with the thing you didn't expect. "We assumed the effect would be strongest in older patients. It was the opposite, and we spent four months trying to explain it away." That's a post people finish reading, because you've told them a story with a tension in it.

End with a real question: has anyone else seen this pattern? What would you check first? You'll get comments from people who've hit the same wall, which is worth far more than likes from people who haven't.

The one-figure post

Pick the single figure that carries your paper. Write two or three sentences on what to look at and why it matters. Don't paste the caption. Captions are written for reviewers; your post is written for a curious colleague on a train.

Visual summaries have their own small literature too (a PubMed search for visual abstracts and dissemination turns up plenty). My own view is simpler: a clear figure with a human sentence beside it beats a designed graphic with a vague one.

The decision you're not sure about

This is my favorite format, and the one researchers are most nervous about. You describe a methodological choice you made, explain your reasoning, and admit where you're unsure. "We handled missing follow-up with multiple imputation, but I keep wondering whether the missingness is tied to the outcome itself. How would you have approached it?"

People worry this looks weak. In my experience it does the opposite. It signals that you think carefully, and it gets methodologists and statisticians to show up in your comments, people who would never respond to an announcement. If you post this while the work is still in progress, you might fix a problem before a reviewer finds it.

The null result

Null results rarely get a press release, and that's exactly why a well-written post about one stands out. Say what you tested, what you found, and what it rules out. Other groups planning a similar study will be grateful, and some will tell you so.

The reading-list post

Share three to five papers that shaped a piece of your work, with one line on what each changed in your thinking. This positions you as someone who reads widely. It also tends to pull in the authors of those papers, which is how some surprisingly good collaborations start.

This is where having search and writing in the same place earns its keep. I'll usually build the list while I'm working in PACR's research search, since the papers I'm citing are already in front of me with my notes on them, instead of reconstructing the list from memory a month later.

Writing the post: a worked example

Here's how I'd approach it with a realistic case. Say you're a clinician-researcher who's just finished a retrospective cohort study on readmission after a common surgery. The headline finding is modest, but you noticed that readmission rates dropped sharply at one site after a discharge-planning change.

1. Write the elevator sentence first. "Readmissions after this procedure barely moved overall, but one hospital cut them sharply after changing how it scheduled follow-up calls." That's your story, and it isn't the one in your abstract.

2. Pick the format. This is a surprise post, with a decision-post flavor, because you're not sure whether the site effect is real or an artifact of who that hospital serves.

3. Draft from your own material. In your PACR workspace, the paper, your reading notes and your summaries of the related studies sit together, so the draft can draw on context you've already built. You aren't starting from a blank box and trying to remember which prior study found something similar.

4. Cut it in half. First drafts of posts are almost always mini-abstracts. Delete the background paragraph. Lead with the site finding.

5. Add the one figure. The site-level plot, not the forest plot of everything.

6. End with a question you genuinely want answered. "Is anyone aware of studies that separated the effect of follow-up calls from differences in patient population?" That's a question a health services researcher can answer in one comment.

The whole thing takes maybe twenty minutes once the habit is in place. Without a connected workspace it takes longer, mostly because you're hunting for the papers and notes you need, which is the part that makes people give up and post a bare link.

Likes are cheap; replies are the point

I'll be blunt about metrics. A post with forty likes and no comments did less for you than a post with six likes and three thoughtful replies. Likes tell you people agreed in passing. Comments tell you someone thought about your work long enough to respond, and those people are your future reviewers, co-authors and critics.

So treat replies as the main event:

• Answer every substantive comment within a day or two, ideally with something specific rather than "thanks!"

• When someone disagrees, ask them to say more before you defend yourself. Half the time they've spotted something real.

• If a comment points you to a paper you haven't read, read it, then say so in the thread. People notice.

• Share other people's posts with a sentence of your own on why they matter. A share with no comment is a like with extra steps.

One more thing on timing: post when the work can still change, not only when it's frozen. The decision post above is far more useful three months before submission than three months after.

What not to post

Being open about work in progress doesn't mean being careless. A few lines I'd never cross:

• Anything that could identify a patient or participant. Case details, rare conditions in small locations, images. The ICMJE recommendations on privacy apply just as much to a post as to a paper.

• Results a journal has under embargo. Check your journal's policy before posting findings from an accepted paper that isn't out yet.

• Claims stronger than your paper's. If the paper says "associated with," the post doesn't get to say "causes." Overstatement in a post is still overstatement, and readers will screenshot it.

• Co-authors' unpublished work without their agreement. Ask first. Disputes about credit start in small moments like this, and COPE's guidance is worth knowing before you need it.

Good reporting habits help here too. If your study type has a reporting guideline on the EQUATOR Network, the items it asks you to report are often exactly the details a sharp commenter will ask about. Having them straight before you post saves you an awkward thread.

And make sure people can find the rest of your work. A linked ORCID iD means that when a post lands well, the curious reader has somewhere to go next.

A second grant officer, a different lesson

A few years after that workshop, I got a call from a program officer named Ruth about a progress report. She'd read it, it was fine, and she had one question: "What did you learn this year that you didn't expect to?"

I didn't have a good answer ready. The report was a list of completed aims. The interesting part, an analysis that had gone sideways and taught us something about our measurement tool, wasn't in it at all, because it didn't fit any aim.

She told me that was the part she wanted. Funders, she said, already assume you'll do what you promised. What they can't predict is what you'll discover, and that's what makes them want to fund you again.

I think about that call whenever I write a post. The completed-aims version of your work is the announcement. The thing you didn't expect is the conversation.

Start with one post this week

Don't overhaul anything. Pick one piece of work, finished or not, and write a surprise post or a decision post about it. Keep it short. End with a question you actually want answered. Then stick around for the replies.

If you're not on PACR yet, you can set up your workspace and write your first post with your papers, notes and drafts all in one place, so sharing what you're working on becomes part of doing the research instead of a chore you tack on afterward.