Before You Build the AI, Ask Who Owns the Number
The Governance Conversation We Needed to Have
Data & AI Stockholm joined forces with Developers Bay at United Spaces Waterfront for an evening on data governance, or as John put it, the most underrated skill in AI. The room reflected it. A genuine mix of backgrounds, industries, and experience levels, some familiar faces and many new ones. One speaker, one topic, and by the end of the evening, no shortage of opinions about both.
Governance is an Operating Model, Not a Policy Folder
John Wernfeldt, Managing Director at Northridge Analytics and President of DAMA Sweden, the Swedish chapter of the global body that literally wrote the data management standard, did not ease into it.
The opening argument was direct: most organizations have confused having governance with doing governance. The policies exist. The frameworks are documented. And yet on any given afternoon, when a real data decision needs to be made, nobody knows who owns it.
His framing recast the whole thing.
Governance is not what lives in a folder. It is what happens between the meetings, when nobody is watching.
John grounded it with an example the room recognized immediately. Sales and finance both report revenue. But do they mean the same thing by it?
Booked or recognized, gross or net, which deals count and from when? When the definitions were never agreed upon, the numbers will never match. And when they do not match, someone needs to explain why. That task lands on a data analyst who did not define either version and owns neither. That is not a reporting problem. That is a governance problem.
And if that is the foundation, the question becomes a straightforward one:
what exactly are you planning to build your AI on top of?
He introduced a framework built around the smallest possible unit of governance:
a metric that matters, a decision it triggers, and one person accountable for whether that decision was right.
The simplicity is intentional. Start there, make it real, then build on it.
The examples he brought grounded it quickly. A bank tracking customer churn with a clear owner for what happens next. A healthcare provider that went from a dormant governance council to an active operating model ahead of a major platform launch. In both cases the shift was the same: from governance as something the organization had on paper, to something it actually did.
The closing argument was one worth taking home.
The race to adopt AI tools is real, but John’s point was that the organizations which come out ahead will not necessarily be the ones who moved fastest on the technology. They will be the ones who sorted out data ownership first.
The Q&A pushed on the human side of this. Audience members wanted real examples: who are the people actually sitting in these rooms, and what does accountability look like when a model gets it wrong. The conversation stayed specific and did not drift into abstraction.
Six Tables, Plenty to Say
Group discussions have become a signature part of how Data & AI Stockholm structures its evenings, and this one was no exception.
Rather than closing on a talk and calling it a night, participants moved straight into six groups to work through the themes from their own organizational experience.
One question that surfaced across several tables was worth sitting with:
what does AI readiness actually mean?
Is it a data quality threshold, an ownership structure, a cultural shift? The room did not land on a single answer, and perhaps that was the point. It is a question worth defining before the work begins.
The debrief was candid. People brought real examples from their own organizations, and the mix of experience levels in the room made for a more honest conversation than usual.
Then the Rooftop
Stockholm delivered. The sun stayed, the rooftop filled up, and the conversations that had started around six tables kept going in smaller groups over drinks. The kind of continuity that does not appear on any agenda but is often what people remember most.
Thank you to Developers Bay for bringing this evening together, to John for a talk that gave a room full of practitioners something genuinely worth disagreeing about, and to everyone who showed up, engaged, and made the evening what it was. The energy in the room was yours.
Data & AI Stockholm 💙
Stay Connected
If you’d like to stay connected with the community and hear about future events, the best places to find us are:
Such great energy in the room, thanks for having me