DPW Amsterdam 2025

At DPW Amsterdam, one moment said it all. A slide with AI use cases went up, and instantly, every phone in the room was raised.

One VP I spoke with mentioned she was looking for a framework to define, rollout, and govern AI projects. She said: "We know we need AI, but where do we even start?"

To cut through the noise, I've synthesized dozens of DPW sessions into a clear 5-step framework for a successful AI rollout:

1. Start with Purpose, Not Tools

Identify real business problems before adopting technology. Avoid chasing “shiny AI” solutions; instead, align every initiative with measurable business outcomes like efficiency, cost control, or supplier optimization.

2. Think Big, Start Small, Scale Fast

Set an ambitious long-term vision (e.g., agentic workflows, AI-enabled orchestration), but begin with manageable pilots. Validate use cases quickly through a “digital garage” or “agent factory” model, then scale successful ones across the organization.

3. Build a Governance & Data Backbone

Establish strong governance, including clear entry/exit criteria for pilots, KPI alignment, and IT oversight. Invest early in data harmonization and ownership - good data management is the foundation for AI maturity.

4. Drive Adoption Through People

Create AI learning journeys, appoint AI ambassadors, and shift KPIs from savings to adoption metrics. Encourage experimentation through hackathons, internal contests, and visible recognition of AI use cases.

5. Balance Quick Wins with Long-Term Integration

Run short-term use cases for momentum but plan simultaneously for deeper integration into ERP and enterprise systems. Combine agility with architectural foresight to ensure scalability and sustainability.

Originally posted by Philipp Gutheim on LinkedIn.