Last month, I attended the Boye and Co Digital Leadership Group meeting at the Royal College of Surgeons in London. The group exists to connect digital leaders across organisations so we can learn from each other’s challenges and avoid mistakes others have already made.

The turnout was smaller than expected. Last-minute cancellations meant the group was more intimate than usual, but that actually made it better. More time for real conversation. Better questions. The kind of peer learning that doesn’t happen in bigger conferences.

Here’s what stuck with me.

On CMS Migrations: You’re Not Alone in This

Sam Hope-Evans at Soundcloud shared the start of their journey moving from a vendor to OpenSource CMS. What struck me wasn’t the technical choice, it was that they were asking the same questions we’ve been asking at Edinburgh: Is our current system giving us what we need? Can we move without losing everything? What does discovery look like when we’re under pressure?

Using AI in their discovery phase to build prototypes and get buy-in from stakeholders quickly without spending valuable time developing prototypes. When they asked me about our Drupal 7 to Drupal 10 migration, I realised how much we’ve learned that’s worth sharing. The pitfalls (data quality, stakeholder fatigue, scope creep). The wins (getting to a modern platform, being able to move faster). The ongoing question they’re wrestling with: cloud-hosted or on-premise?

If you’re thinking about a CMS migration, the lesson is this: talk to people who’ve done it. Not vendor reps. People in organisations like yours who’ve lived through it. That peer learning is invaluable.

On Legacy Systems: The Long Game

Mahesh Patel at The Royal College of Surgeons shared how they consolidated from 100 separate CMS instances down to one Umbraco system. They’re also grappling with medical data systems that need overhauling. How do you start looking at legacy infrastructure when you’re managing multiple competing priorities?

This resonated deeply. We’re managing four services simultaneously across the university. Our website platform EdWeb 2 was a significant architectural shift: we rebuilt from Drupal 7 (which used organic groups to simulate subsites within a single installation) into Drupal 10 with 160+ separate websites built on a single codebase. When we push updates, all 160 sites get them automatically. It’s modern infrastructure, but we’re still evolving it.

MyEd is different. We’re managing the legacy system in production while simultaneously starting a parallel modernisation project to rebuild it from the ground up with Drupal at the heart. That’s not a simple upgrade. It’s building the new while keeping the old running. Staff Profiles is coming next. The question of ‘how do we tackle these transformations without breaking what’s already working?’ is one I’ll been trying to figure out for years.

Their approach resonated with me: be clear about what needs to move first, understand the data, plan for the long game. Don’t try to solve everything at once. That’s exactly what we’re doing: one service transformation at a time, while maintaining the others.

On AI: The Uncomfortable Conversations

Simon Jones at Studio 24’s presentation on the ethics of AI pulled no punches. And it’s worth repeating what they said, because it challenges the hype.

AI is being sold as a solution to everything. But the reality is more complicated:

The real costs are often invisible:

• Financial: AI companies are running at a loss. Most aren’t profitable. You’re investing in something that hasn’t proven it’s sustainable and it’s becoming more expensive than developers.
• Human: Data taggers in developing countries get paid little wages. Communities live near data centres can strain local resources (AI consumed 21% of all electricity in Ireland in 2023). The environmental cost still unknown.
• Creative: Copyright issues (Authors suing OpenAI over copyright infringement (https://www.bbc.co.uk/news/technology-66866577). Bias in training data and Quality dropping as AI trains on itself.

The harms are real:

• Phishing attacks and security risks are on the increase.
• Misinformation and ‘AI slop’ being generated which users are believing.
• Bias in training data and system prompts.
• Filter bubbles and reduced trust – AI starts to mimic the user.

But there are genuine opportunities:

• Understanding large codebases and content
• Speeding up routine tasks
• Research and discovery
• Accessibility (tools available like ai that focus on inclusive models)

The key question Simon posed: Before you adopt AI anywhere, ask yourself:

1. What’s the business value?
2. Is it high-risk? (Chatbots, for instance, can be high-risk and not actually helpful.)
3. How important is accuracy?
4. Who benefits? Who could be harmed?
5. How can you test it?
6. If it goes wrong, how do you give people recourse?

This matters because AI FOMO is not a strategy. Just because everyone’s talking about it doesn’t mean it’s right for your specific problem.

What I’m Taking Away

Three things:

1. Peer learning beats everything else. The group exists because we’re all managing similar challenges in different contexts. The value isn’t in vendor pitches or abstract frameworks. It’s in hearing from the Royal College of Surgeons about consolidating systems, from Soundcloud about CMS choice, from Studio 24 about navigating AI responsibly. You can’t get that in a conference. You get it from peers who are living it.
2. The unglamorous work matters most. CMS migrations, legacy system consolidation, data quality. These aren’t flashy. They don’t make headlines. But they’re the foundation of everything else. The organisations doing well aren’t the ones chasing every new trend. They’re the ones making deliberate choices about their platform and sticking with them.
3. Be cautious of cost, all kinds of cost. Financial, human, environmental, security. When you’re evaluating any new technology (AI included), understand the full picture. Ask the hard questions. Don’t let FOMO drive the decision.

Should You Attend Next Year?

Even with lower attendance, this was worth my time. I came away with real insights, real connections, and a clearer sense of how other digital leaders in the UK are navigating the same pressures we are.

If you’re leading digital work in a university or similar organisation, find peer groups like this.