Another tracker won't fix weak reporting

Weak reporting is often mistaken for a visibility problem. The response is predictable: build another tracker, assign someone to maintain it, schedule another monthly review. Yet these measures do little to address the underlying issue. A tracker can only reflect the information people choose, or remember, to enter. When data is being gathered weeks or months after events have occurred, recollection takes the place of evidence. On paper, the reporting process may appear comprehensive. In reality, it can be capturing very little, timely or reliable.

Through my own journey at Researchfish, here’s what I learnt to make reporting and tracking quicker and smoother:

1. Start with the information you're always chasing

Go back to your last reporting period. What did you have to ask for twice? What did people end up digging through old emails, calendars and meeting notes to find? What depended on someone remembering something that happened months earlier?

Be precise about this. "Engagement activity" isn't a useful requirement. What you actually need is the name of the event, the date, the audience, the purpose and the project output it supported. Vague requirements get you vague information back. People give you what you ask for, so ask for exactly what you need.

2. Find where that data already exists

Most of the evidence you end up chasing wasn't generated during the reporting process. It was generated earlier, when the work was being organised or signed off. Someone attending a conference needed travel approval. A team planning a workshop needed a budget signed off. A supplier didn't get paid until the work was confirmed complete. These processes are reliable because the activity can't continue without them. A separate tracker rarely carries that same weight. The information is already changing hands somewhere in your organisation. Find that point, and that's where you collect it.

3. Ask for the minimum you'll need later

Don't turn an approval form into a second report. Ask only for what you'll genuinely need down the line. A short requirement attached to a decision someone is already making is more likely to be completed properly because the context is still clear. If the conference name is already on the travel approval, you know who's going where, on what date, for what reason and against which project. That's easier to capture in the moment than to reconstruct at year end, when everyone's buried in their own reviews.

The point isn't to add more process; it's to capture a small piece of useful information while it's still fresh, using a step people are already taking.

4. Check whether the chasing has actually stopped

Next reporting cycle, don't just check that the new form got filled in. Check whether fewer reminders were needed. Check whether the information came back more complete, and whether the report took less time to put together. If the same gaps show up again, the requirement has probably been attached to the wrong process, or it's asking for more than people can reasonably give at that point. The real test isn't whether a new field exists. It's whether reporting has become part of the work itself, rather than something reconstructed afterwards.

How it worked in practice

I used this approach while managing reporting for the Data and Digital workstream of a nationally funded UK research programme at Swansea University. Each year, the team spent one to two months preparing its Researchfish submission. Researchfish is used by many UK research funders to collect information about research outputs, outcomes and impacts. The work had been done. There just wasn't a reliable record of it anywhere. Information was scattered across emails, calendars, expense records and people's memories, and preparing the submission meant the administrator had to chase down researchers individually to reconstruct a year's worth of activity.

I spoke separately with the department head, senior researchers and the administrator handling the submission to understand where the process was breaking down. Most of the missing information turned out to be linked to activities that already required approval. Conference travel, workshops and events all went through an existing sign-off process before the related costs were approved.

So I built a short form based on the relevant Researchfish fields and attached it to that approval process. The information had to be entered before the associated cost was approved. The next annual submission was ready in under a week, preparation time fell by more than 70%, and the administrator stopped spending weeks chasing researchers for details about events that had happened months before. No new platform. No new tracker. Just a change in when the evidence got captured, at the point it already existed rather than months after the fact.

Conclusion

None of this requires a new system, increased manpower, or another Monday meeting. Before looking for a tool to track down information again, one needs to first find out where it currently exists. The majority of reporting issues are not due to a lack of data, rather, they stem from timing issues, with evidence collected too late to be reliable and too disconnected from the task to be easily found. If you fix the timing, the tracker will be redundant.

Don't begin by creating something new if you are in charge of reporting on a funded programme. To begin, find the location where your evidence is already created, identify the approval step that is closest to it and put your capture there. That's the entire process. It works because it fits how they now perform their jobs instead of asking anyone to take on an additional step.

Partly AI supported; human-validated