Data is not just a by-product of operations – it is a strategic asset. Our recent panel discussion at HEPN North explored how institutions are using data to drive smarter decisions, improve engagement, and achieve academic and operational gains.

During the discussion, we heard from Matthew Holden, Head of Educational Technology at the University of Salford and Laurence Eagle, Independent Consultant at Advance HE who shared practical examples from across the sector, showing how universities are overcoming silos, embedding analytics into everyday practice, and using insight to fuel performance, accountability, and institutional resilience.

Here are some of the key takeaways…

From overload to insight: tackling the university data challenge

Universities collect massive amounts of data from student records, learning records and support systems. The problem isn’t the absence of data – it’s the inability to access, trust, or effectively use it. So, despite being ‘data rich’, many universities struggle to make proper use of what they have – leaving them ‘insight poor’.

A major contribution to this insight gap is the fragmentation of data. As Matthew explained:

‘Universities have lots and lots of systems. We need to reduce the amount of systems and we need to be clear of what the purpose of each system is… Master the data once on one system and move or integrate that between the systems.’

Without integrated systems working as one, we’re left with fragmented data, a poor student experience and a lack of trust in the information we rely on. As Matthew said:

‘We need to get to a point where we can trust our data and have it readily accessible. If we can manage that, that is the start of a journey. But until we get that data right, it’s a difficult journey to start.’

Dashboards, trust, and data ownership

Good data is fundamental when it comes to making well-informed decisions, so it’s important that we’re able to trust the data that is presented to us. While dashboards and tools like Power Bi can help us make sense of large amounts of data, they can sometimes expose flaws in underlying information.

While some people might see this as a setback, Matthew argued that exposing data flaws is actually a step forward, stating, ‘If we’re collecting the right data and it’s wrong, you shouldn’t be afraid of surfacing that because if we don’t surface it, it’s not going to get corrected.’

In order to make sure they’re getting the right information; organisations need to take ownership of their data. As Laurence pointed out, ‘You have to have somewhere at the end user, checking that it is right. If you put this data out, the person needs to take responsibility to say this is correct.’

Continuing the conversation on ownership, Laurence spoke about a time where he had to challenge incorrect reports until the data was finally accurate.

‘We spent weeks going round and round until eventually we got to a set of data that we said, right, this is now correct. So, I was confident in having a discussion then, but it did take a quite a few iterations before we got to there. But at least I felt like I was involved in the process.’

Laurence also discussed positive uses of dashboards, reflecting on a time when Roehampton University used a traffic light system to target support for students most at risk.

‘It took a lot of work to make that dashboard that simple. It’s not simple, but it looks simple from the end users’ point of view. So, I thought that was quite an interesting use of visualisation of data.’

This approach demonstrates how dashboards, when built on clean and trusted data, can drive timely, human-centred action.

Beyond the numbers: toward a data-informed, human-centred university

While ‘data-driven’ has become a popular term in higher education, Matthew challenged this, questioning whether it truly captures the relationship universities should have with data. He explained, ‘I don’t like the term data-driven because data-driven sounds like you’re just using the data to make a decision. So maybe something more like data led or data informed.’

Lawrence then shared a past experience leading a cross-university scheme. With no clear performance benchmarks, he built his own dataset and analysis tools. He explained:

‘I gave them a target, so they had something to discuss, something to aim for, and something to make action plans about it. But I had to get the data clean which took me three days but also have an intent. And I think what you were saying about data-driven, what’s the intent behind us collecting this data? What are we using it for? And I think it’s things like that that we also have to think about when we talk about data.’

Instead of focusing solely on numbers, Matthew argued for a more human-centred approach when interpreting data, especially when it comes to using AI. ‘We want to use the data, but there’s got to be a human element of how we interpret some of that in terms of AI.’ He continued:

‘If you put AI on top of poor data, it’s going to give you the poor answer. I think that’s why we need to look at the basics. Let’s get some real accurate data and then start putting the AI on top of that. It’s going to take time, but if you can segment your verified accurate data and just tap your AI into that and start moving your existing potentially poor data over in time, it means that you can start leveraging some of the benefits of AI but with accurate defined data set.’

Join our future discussions

A big thank you to our speakers Matthew and Laurence for sharing their insights. If you found this conversation valuable, we’d love to see you at our next HEPN event, where we’ll continue exploring the big questions shaping the future of higher education. register your interest to join us at HEPN South on 13-14 November 2025!