The Architecture of Trust and Experience: what 40 data leaders told us about where Belgian organisations actually stand

The numbers don't lie. The strategy documents don't exist.

Siegert Dierickx (Co-Founder and Managing Partner MultiMinds) opened with the benchmark findings, and the first number set the tone for the whole evening.

90% of marketing and data leaders say management believes in the power of data. Only 33% have a documented data strategy.

That gap between conviction and commitment is where most organisations are stuck. Belief is cheap. A documented strategy means someone has signed off on priorities, tradeoffs, and accountability. Without it, data initiatives live and die by whoever is most persuasive in the room that week.

The AI numbers were just as revealing. Every respondent uses AI tools at least weekly. 86% use them daily. Yet only 19% say they understand AI's impact on their business "very well". That's not an adoption problem, Belgian organisations are using the tools. And that is harder to fix.

On integration: respondents pull from an average of six data sources, but fewer than half can combine them into a single customer view. And activation (actually using collected data to drive decisions) scored an average of 54 out of 100. The weakest dimension across the board.

Data is being collected faster than it is being used. That's not a storage problem. That's a priorities problem.

What closing the gap actually looks like

The evening's keynote came from Aline Cadron, Product owner Data Insight at VRT, who brought something the benchmark couldn't: eight years of experience doing the slow, unglamorous work of building data culture inside a large organisation.

She introduced a concept worth holding onto: algotorial - the space where editorial judgment and algorithmic personalisation are designed to work together, rather than fight for control. For any organisation where humans still make meaningful decisions alongside automated systems, that tension is live. Getting it right is design work, not configuration work.

Her closing line was the sharpest thing said all evening: "Governance is a process, not a one-time project."
If your organisation finished a governance initiative and moved on, you didn't finish the project. You paused it.

The panel: three perspectives, one honest conversation

The evening closed with a panel that covered ground most conference panels avoid.

Aline Cadron (VRT) held the practitioner's ground. Eight years in, she knows what it takes to move a large organisation from data ambition to data behaviour. The algotorial concept she introduced in her keynote became a thread in the panel too: the question isn't whether humans or algorithms decide, but whether you've designed the relationship between them deliberately. Most organisations haven't.

Bart Van den Brande (Sirius Legal) made a point that reframes how compliance usually gets discussed: privacy is no longer a policy document, it's an architectural decision. If your data infrastructure isn't built with privacy in mind from the start, no policy will save you. This has direct implications for how data teams scope their work, not just how legal reviews it.

Marc Decorte brought the boardroom view: data and AI investments compete for executive attention the same way everything else does. Most don't survive that conversation; not because the business case is wrong, but because the people making it don't speak the language of the people approving it.

Philippe Vlaemminck closed with the number from our benchmark that should follow every one of our clients into their next strategy review: the biggest gap between data leaders and data laggards wasn't AI capability, budget, or strategy sophistication.

It was execution: a 67-point gap.
Everything else is a proxy for whether your organisation actually does what it says it will do with data.

What this means

The organisations in the room were mid-sized to large, Belgium-headquartered, and serious about data. They weren't at the start of their journeys. Most had already invested in tools, run pilots, and stood up data teams.
And still: two-thirds had no documented strategy. Half couldn't join their own data sources. Activation was an afterthought.

If that sounds familiar, the problem isn't your tooling, it's the layer between what you've built and how it gets used.

We run this event every year because we think these conversations matter. It is the only honest starting point for the work that follows.

If you'd like to read the full findings from the MultiMinds Marketing & AI Maturity Benchmark, get in touch and we'll send it over.