Social media posts about research
A social media post about research has to earn a reader who can scroll past it. Writing one turns on choosing the platform and the reader, a first line that earns attention, how much that line gives away, the structure and wording of the post a journal asks for at submission, and stating how strong the evidence is in words a lay reader will not misread.
Platform and audience
A post on social media is not written the same way for a scientist, a doctor, a patient and the general public. Social media is full of people, so writing for it comes down to understanding whom the post is meant for and what that reader wants from it.
In April 2026 most medical writing on social media went to Twitter/X or Blue Sky, which carry short posts that can run on as threads. Twitter/X suited storytelling across a thread with a link out to the full content, and it reached a very large general audience. LinkedIn took longer posts in a more professional tone, with room for a lot of data. Its audience was more professional, with more journalists, investors and doctors, and usually wanted leadership insights or takeaways, but its algorithm was very odd, and a post there usually reached fewer people. Instagram was visual and casual, built on infographics, video, highlights and visual storytelling; most writers made little use of it, though it was a good place to communicate.
On social media the visual is sometimes the primary medium, so an image that carries a post should be self-explanatory rather than described in the text. Some algorithms, LinkedIn's among them, promote a post more when the image says something different from the text.
The hook
The most important thing in writing a social media post is to earn the reader's attention very fast, because social media holds tens of thousands of people who will scroll past a post, and the hook, the first thing a reader sees, is where that is done. Without a hook, readers have very few reasons to keep reading. A hook should be relevant to the reader and timely, and it should not be salesy: marketing is acceptable, but a reader who immediately thinks they are being sold to will not engage. Emotion is one big difference between social media and most medical and scientific writing. Scientists are used to rigour and data, not to the human side, and emotion should not be used alone to communicate scientific data, but emotion in hooks and posts makes people want to read more, because people are reading them.
A good hook has as many as possible of these: personal experience, which signals that a subject relevant to the writer may be relevant to the reader too; a question, which makes the post about the reader and makes them stop and think; emotional relevance; visible expertise; and the topic itself, since a line as broad as "something happened to me yesterday" gives nobody a reason to stop. Curiosity, humour and data are very good ways to interest a reader: something funny or something that makes them curious stops them scrolling, and a figure such as 90% of people suffering from something makes readers likely to think they are among the 90%. Visuals help where they are not too dense. Combining several types gives the best hook, though it can end in a very long one.
Hooks written for one treatment against migraine show four ways to open. As a question: "What if you could cut your migraine days in half? A new study shows this may be possible for millions of people", which says whom the post is for, prompts reflection, appeals to emotion and signals new data. As a statistic: "Seventy percent of patients in a major study reduced their migraine frequency by 50% or more", which lets a reader decide at once whether they are interested. As emotion: "Imagine getting back four days every month. Four days without pain. Four days to live your life. And for millions with migraines, this just became a real possibility.", which may compel a reader who has known pain to click and read on. As journalism, for an audience of journalists, in big bold letters: "Breakthrough: New treatment changes everything for migraine sufferers", which is again very emotional. A combined hook, a little longer, opens "What if you could get back four days every month?", breaks up the pattern of the text with an emoji, and continues "67% of patients in a new study got exactly that. Here's how this migraine breakthrough is changing lives.", carrying relevance, emotion and a question.
How much the hook gives away
A post needs only enough to catch a reader's interest, and trying to fit the whole story into it is a classic scientist's mistake. Every word of a hook qualifies the reader, telling them whether the post is for them or not, so the more words a hook has, the less likely a reader is to click.
The length of the hook is a trade. A hook cut down to "What if you could have four days every month without pain?" could be about migraine, a period, broken bones or chronic pain, so more people click and read; the cost is that once they see it is about migraine, only a subset of those who clicked keep reading. The choice depends on whether the post seeks broad attention or the attention of one group of people. Curiosity works really well, and a hook with as little in it as possible, so that whoever clicks will at least enjoy it, makes a very good curiosity hook.
The post that accompanies a publication
In health research social media had become a tool to recruit patients, collect data, develop engagement and disseminate research, particularly for researchers concerned with demonstrating the impact of their work, although it was not yet clear whether posts increase citations. It was important where the data had public health implications, may influence policy-making, may improve health systems or have a health-related socioeconomic impact, and publication writers were increasingly faced with producing research-based copy for it, a change that medical writers, as knowledgeable and flexible science communicators, were well placed to support. The tools that turn a clinical study report into an article are the ones a post needs. As the last link between the author, who knows the data intimately, and the reader who will discover it, it falls to the publication writer to carry the job a step further and promote the value of the research.
By December 2021 a journal's submission system could end with a box inviting an optional Twitter post text for the manuscript, and journals asked authors for a single post to promote their work.
The word tweet was coined for a short burst of inconsequential information in only 140 characters, and in December 2021 a tweet allowed 280. Squeezing a 250-word abstract into one means writing very concisely and precisely, in simple sentences, and sticking to a structure, because a post has an established structure just as a publication does. The first line is the catchphrase or title, which usually appears in notification pop-ups on smartphones and computers and is the place to capture the reader's attention. Good ways to do that are to ask a question, state a fact, state the context or the general problem in the domain the study addresses, highlight the problem the study addresses, or place a key, attractive, alarming or eye-catching figure. The text after it is a concise storyline that summarises the study's conclusion, perhaps with a fact, some supporting data, or information in decreasing order of importance. Its sole aim is to entice the reader to click the link, read more or share the key data.
Because each character counts, readable, concise and precise writing is paramount, and five writing tips serve it:
- Choose active verbs.
- Put the most important verb idea early in the phrase.
- Make the vocabulary reflect the context, methods and data accurately.
- Consider presenting the study design as PICO: population, intervention, comparator and outcome.
- Place the key result data with their significance. Superlatives such as superior or better, and verbs such as improves, increases, decreases, faster or slower, need comparative data with appropriate significance behind them.
Hashtags and tags
Hashtags are keywords that let communities interested in the research domain engage with the content, as #Alzheimers may be followed by anyone interested in the disease, its management and its treatment. They usually sit at the end of the text, inside the 280-character count. Most often the right one already exists, and it can be found by searching the research theme on Twitter, with tools such as Tweetdeck or Google, or on the websites of organisations and companies. A new hashtag is worth creating only when the subject is really new, as COVID-19 was in 2019, or to attract a specific audience, such as a French-language Alzheimer's tag for readers in France.
Tags address specific accounts of individuals and organisations, such as @Alzheimer's France, the first or corresponding author or another expert who supports the publication, so that they retweet, comment or add information, opinion or support. Where the writer is unsure of an expert's position or reaction, a private conversation with them by direct message gives the expert the chance to engage without doing so publicly. Tags can also go in the legend of the central illustration, where they do not count against the characters.
The central illustration
The central illustration completes the story as bonus content outside the character count. It is an image, a figure, a graph, or a link to a webpage or a PDF that supports the post's claim with complementary information summarising or highlighting the data, and it is a further place for tags to authors, organisations or associations concerned with the problem.
Threads
Posts can also be planned. Beyond the single post a journal asks for, some organisations, patient associations or authors post a thread, a Twitter plan of usually up to 10 posts, each about one part of the research story taken from any part of the manuscript from background to discussion. Posts on the discussion points are especially valuable, because they can draw other experts into discussing the data or its implications. A medical writer can offer such planning to clients as an additional service.
Evidence strength in words a lay reader will not misread
The internet has become the main source of health information for many patients, but most health information on social networks is unregulated and of questionable quality. Unlike sponsored medical communication and health journalism, blog posts and personal commentary channels have been completely unregulated, and influencers gather large numbers of followers who pay close attention to their views and are guided by their choices. Scientific publication is usually regulated by peer review, but for the general public that check does not exist, and a press release, a blog or a simple statement citing data can sometimes confuse and raise unfulfillable expectations.
With the amount and spread of misinformation during the COVID-19 pandemic, Facebook and YouTube became acutely aware of the problems of social media as a channel for health research. Facebook's head of health proposed that the long-term solution is to connect people to accessible, trusted information, and YouTube's global head of healthcare and public health spoke of plucking the weeds of misinformation and replacing them with good, engaging information.
A writer reporting a study to a lay audience should say whether it was peer-reviewed and explain that this means other experts in the field checked the validity and quality of its reporting. The writer should also place the data in the long process of building evidence, as in "this is the first study in humans and the results will now need to be confirmed in larger, controlled studies", so that the public knows what to expect and how to act.
Each word has to reflect the data accurately. "This study shows treatment X is effective" and "this study suggests treatment X is effective" may look the same to the public, and so raise a false certainty that is relayed across social media platforms. Significantly has a specific scientific meaning and should be used only where a p-value is associated with the data point. A non-significant result does not mean there is no clinical difference, since the sample size and other factors decide practical significance, so the writer should place the result in context.
Short sentences, a straightforward message in a relaxed tone and everyday language suit the lay reader, and bullet points make complicated concepts easier to capture, understand and remember. Imagery, visualisation or numeracy are useful ways to reconceptualise health data, as the WHO's infographics on Facebook did during the COVID-19 pandemic. Used responsibly, social media could be a useful tool to educate patients and promote public health messages.