Research centres generate a constant stream of material worth sharing: findings, commentary on developments in the field, reflections from events and conferences. The challenge is rarely a shortage of things to say. It’s getting that material out of researchers’ heads and onto a page, in a way that fits around everything else competing for their time.
I’m Jack, a web officer within a research team. My work centres on increasing the visibility of that team’s output – increasing readership, engaging funders, and giving people more reason to keep following what’s happening. All of which depends on solving exactly that problem.
The result is a short Microsoft Forms tool called the 5-Minute Check-In, developed through the TIDE project and tested with DDRC, designed to turn a few minutes of a researcher’s time into a polished blog post or LinkedIn update.
Designing around what researchers will use
The starting point was a conversation with the project sponsor and other stakeholders, who were also the people the framework needed to work for. The aim was to find out what they would realistically be willing to engage with, rather than making assumptions.
That conversation set the direction for the whole project: an open-ended brief to “write a blog post” was never going to get much take-up. What worked better was a short, structured prompt that researchers could answer in a few minutes, with the heavier work of turning that into a draft handled elsewhere. Keeping the thinking with the researcher, and taking the writing effort away from them became the organising principle behind everything that followed – right down to the form itself promising a five-minute time commitment and explicitly inviting people to answer “even one or two” questions rather than all of them.
One form, not three
The 5-Minute Check-In is built around three types of content researchers typically have to share: a research outcome, commentary on a recent development in their field, and reflections from an event. Each type has its own set of questions, designed to draw out just enough material to feed into an AI drafting prompt without asking anyone to produce a finished piece of writing.
Initially, I grouped these into three separate forms by category. However, testing with a small group of researchers surfaced two pieces of feedback.
- First, people weren’t always sure which form their update fitted into; a unified single form took away the guesswork. The form now opens with a single dropdown question – “What type of research output?” – which branches respondents straight to the relevant section, so no one needs hunt for the right form.
- Second, some updates didn’t fit neatly into any of the three categories, so a fourth “Other” section was added, asking the same shape of question – what would you call this, what’s the overview, what’s the key takeaway – so it still produces something usable, just without forcing a format.
The Research Insight section also asks respondents to flag whether the work is early-stage, in progress, or published – which helps me see whether to hold fire or proceed down the pipeline.
Timing the follow-up automatically
Reflections on an event are often more useful a few days after it, once there’s been time to process, rather than immediately. To capture that, a Power Automate flow reads the event date from the original submission, adds a set interval, and automatically sends a short follow-up form once that time has passed – removing the need to manually track and chase individual events.
The Event Promotion section also asks for images and the correct way to credit them at the point of submission, so that detail doesn’t need to be chased separately once a piece is ready to publish.
AI as a starting point, researchers as the final word
Once responses come in, they’re fed into an AI prompt requesting a draft blog post in the tone of an academic researcher in the relevant discipline, written to read credible and measured rather than promotional. The researcher then adjusts the draft until it sounds like them, and the same prompt can be extended to generate social media posts alongside the blog draft, so one set of responses can produce more than one piece of content.
Where the framework sits today
Running this pipeline still depends on a Professional Services role (like mine) to coordinate it: deciding which researchers to approach, chasing responses, feeding answers into the AI prompt where that isn’t yet automated, reviewing drafts before they go back to researchers, and deciding whether each finished piece goes straight to social media or into a content calendar for later.
Further automation of the AI drafting step is possible through Power Automate, depending on the AI credits available within that environment, and is likely the next stage of development. Even then, coordinating the process – deciding who to approach and what’s worth publishing – is likely to remain a role for a person rather than a workflow. But the benefit is that it frames and funnels an amorphous process into something that feels more achievable for everyone involved.
What the framework offers, even in its current form, is a structure that other research centres facing the same challenge could adapt: a way of turning researcher input into publishable content on a predictable rhythm, without placing the burden of writing on researchers themselves.
If your research group has someone who could take on the coordinating Professional Services role, this framework is designed to be copied rather than reinvented – feel free to adapt the form and the drafting prompt for your own team.
If you want to view the form yourself, the link is below – though note it’s built around DDRC’s workflow, so treat it as a starting template rather than a plug-and-play tool. And if you’d like a hand setting something similar up, that’s a conversation worth having – get in touch: j.ryan-phillips@exeter.ac.uk
View the 5-Minute Check-In form here.