#23 (EN): Where Is Your Supermarket Checkout?

Daniel and Nova talk about what gets lost when IT leaders consolidate platforms without ever sitting down with the people who actually use them. The episode takes its cue from a supermarket chain that sends its executives to work a shift at the checkout, and asks what the equivalent would look like in IT leadership. Daniel reflects on platform consolidation, the migration of business unit and shadow platforms into a shared solution, and why the human history behind self-built tools deserves more attention than it usually gets. He names three things that get overrated in consolidation projects (top-down mandates, requirements workshops, and feature parity) and argues for watching real work happen instead, borrowing ideas like paved roads, Gemba Walks, and Genchi Genbutsu. They also discuss which metrics actually reflect user experience during a migration, how AI can help analyze tickets and feedback at scale without replacing direct observation, and Daniel's own gaps as a leader who mostly hears about users secondhand. He commits to two concrete steps before year end: talking to business units about why they built their own platforms, and spending an intern day with a customer team. Key topics: - Why platform consolidation projects often treat business units mainly as migration objects instead of understanding their actual needs - The hidden knowledge built into self-built departmental and shadow platforms, and why shutting them down without analysis causes the same problem to resurface - Three overrated tools in standardization projects: top-down mandates, requirements workshops, and demands for full feature parity - The concept of paved roads or golden paths as a way to make a shared platform genuinely attractive rather than just mandatory - A four-level closeness pyramid for understanding users: reading reports, asking directly, observing work, and doing the work yourself - Which metrics reveal real user experience during migration, such as time to first production deployment and average ticket wait time - How AI can analyze large volumes of ticket and feedback data, and why that still cannot replace direct observation of daily work - Daniel's concrete commitments for the next twelve weeks: talking to two business units about their self-built platforms and doing an intern day with a customer team Key takeaway: Standardizing a platform also means promising people they will be able to work well with it, and that promise has to be checked against reality, not just against a board resolution. Reports and surveys only get you to the first rungs of understanding users. Real insight comes from watching people work, or doing the work yourself.