Freedom, Chaos, and the Need for Governance

During a recent business trip to Edinburgh, I explored a number of Scotland’s castles with my family who had accompanied me. Craigmillar Castle was one of our favourites to explore and tucked inside the gift shop between novelty plastic swords (which my son had a tantrum about not getting) and tartan bookmarks, we picked up a children’s book titled “William Wallace and the Battle to Free Scotland.” It became an instant hit with the kids. That night, I read it to them in my best Scottish accent (which sounds exactly like my Irish accent, but louder). The kids were enthralled. Their favourite part of the story was when William stood up to the sinister Earl of Surry before the Battle of Stirling Castle, challenging the Earl to “meet him beard-to-beard!”.

They loved the idea of someone standing up for what’s right, fighting for freedom against overwhelming odds. It was stirring stuff. As a BI consultant, I’ll admit my mind wandered in a slightly strange direction.

You see, freedom sounds great. It makes for excellent movie speeches and gift shop literature. But anyone who’s managed a shared Power BI environment knows that too much freedom, the kind without any structure, doesn't look like a charge into battle. It looks like 50+ reports, six different filters for the same column, and about 30 visualisations, mostly matrix/table, on a single report page that freezes every time someone opens it on a Monday.

‘The quest for freedom in BI’ image thoughtfully produced for Ascendant Analytics using Canva©

Freedom, in business intelligence, is complicated.

We’re living in an age where data democratisation is the ideal. Everyone, from interns to executives, can build their own reports and explore the company’s data. We’ve moved away from the gatekeeping days of the centralised reporting team and we celebrate empowerment. We hand out Power BI licenses like free samples. Then we act surprised when the Premium capacity starts straining under the weight of all this “liberation”.

The problem isn’t that people are misbehaving. Most analysts aren’t trying to crash the capacity. They’re trying to answer questions. They’re trying to help their teams. Just like Wallace, they’re working towards something that matters to them.

Unfortunately, when everyone’s fighting for their own version of what matters, using a shared semantic model and shared resources, things get out of hand quickly. One team’s urgent board pack ends up slowing down another team’s operational dashboard. Someone adds a 100,000-row matrix visual because their manager “likes to see the raw data,” and suddenly every report in the Fabric capacity feels like it’s trying to load in slow motion.

In moments like these, it’s tempting to respond with a crackdown. Lock everything down, strip permissions, revert to a world where no one can publish unless they’ve completed a ten-step governance ritual under a full moon. The trouble is, that’s not freedom either.

Later in our trip we had the extreme pleasure of stumbling upon a Medieval Fair Day, complete with a jousting show, while visiting Linlithgow Palace. The kids were beside themselves with excitement, and so was I, but as I watched the knights competing in the tournament I realised something: all societies need rules. Not the kind that stifle. The kind that make freedom possible. Wallace wasn’t fighting for total anarchy. He was fighting for a nation’s right to govern itself wisely. And for people to shape their own destiny, within boundaries that protect their collective well-being.

It’s the same in BI. If we want teams to explore data, ask their own questions, and build reports that matter, we need to create conditions that support that freedom. We need boundaries that protect the commons. We need guidelines that keep the semantic model healthy and the shared capacity responsive.

This series explores how to do exactly that.

We’ll look at how Systems Thinking, popularised by Donella Meadows, can help us govern without bottlenecking. We’ll dig into practical guidelines for report builders. We’ll explore how incentives, feedback loops, and light-touch governance can support healthy autonomy. And we’ll do it without resorting to medieval punishments. Mostly.

Because freedom in BI doesn’t mean doing whatever we want with the data. It means doing something meaningful, together, without burning the whole platform to the ground.

So sharpen your dashboard blades, the battle for sustainable BI freedom has begun.