Many data & AI conversations start with technology. Tobias Riedner starts further back, with the business strategy, the stakeholders, and the impact you can actually measure.
He spent 4.5 years as Head of Data Analytics & AI at Danone. Now, as Vice President Data & AI Strategy & Culture at BARC, he draws on that hands-on experience and shares four lessons every data & AI leader should know.
1. The business goal sets the agenda
“There is no data strategy, only a business strategy. Data and AI capabilities have to support it as much as possible.”
The business strategy defines what data & AI should achieve. It sets the goal, whether that’s growing revenue, serving customers better, improving product availability, or something else.
That’s where the real job for data & AI begins: making the right data available at the right time and turning it into usable reporting or AI capabilities.
Once you frame it that way, platform user counts and the number of data products stop being the point. Those numbers can help, but they miss the point. Real impact shows up when data & AI feeds directly into the goals of the business.
2. Lead with technology and you lose the room
“In the beginning, I talked a lot about data platforms and technology. That cost me many stakeholders.”
Lead with platforms and technology too early, and you risk losing people along the way. What matters is early alignment. Before anything starts, the right stakeholders need to know where the initiative is headed and what data & AI will do for their part of the business. That shared direction is what eventually produces fast, effective solutions.
3. Closer to the business, faster and cheaper
“Bringing the data team in-house was exactly the right call. After that, we were three times cheaper and five times faster.”
The lesson isn’t just insourcing. What counts is whether data & AI sits close enough to the business questions to understand requirements quickly and ship solutions without long back-and-forth.
For more on building data capabilities from within, see Stop Buying Data Talent. Your Next Data Leader Is Already on the Payroll. The Merck example shows why upskilling people in context beats buying standard training off the shelf.
4. No democratization without governance
“Data democratization is important to empower employees. Data governance sets the rules of the game.”
Democratization gets misread a lot. It doesn’t mean everyone can do anything with data. It means people can use data sensibly to do their jobs better.
That takes clear rules. Who can work with which data? Which data is cleared for which purposes? Where do you need sign-off, training, or technical guardrails? These questions become more pressing the moment more people start working with data and AI.
Governance doesn’t slow democratization down. It’s what makes it scalable. Without it, democratization stays too cautious to be useful or too risky to trust in production.
The common thread: stay close to the decisions
All four lessons point to the same thing. Data & AI doesn’t create impact through more platforms, more data products, or more use cases. It creates impact when it sits close to the decisions that move the business.
The BARC Data, BI & Analytics Trend Monitor 2026 confirms the pattern: across 1,579 professionals, data quality, culture, governance, and data & AI literacy all rank ahead of AI and ML. Maturity rests on organization, roles, and collaboration, not on technology alone.
For data & AI leaders, that means talking less about technology as an end in itself, and more about how it leads to better decisions, faster solutions, and safer use of data.
