Four Months of Build Days: What Vandalf Taught Me About Program Management

I built a camper van over four months — twenty-plus build days, one system at a time, every piece of it photographed and argued over. At the end I walked through it on camera. It felt exactly like walking around my own decisions.

The bench that became a cabinet that became a bed. The shower base test-fit into the layout before the walls closed up. The water tanks and sink that turned a cargo van into a camper. Everything got built in the shop first, then argued with, then rebuilt.

Here's what the build taught me that I use on every program since: sequence by dependency, not by enthusiasm.

The bed went in first, not because it was the most fun, but because it was the anchor. When the bed's in the right spot, everything else falls in around it. Get it wrong and you're redoing the layout for the rest of the build. Every program has an anchor like that — the integration that everything else hangs off. Most teams start with the exciting pieces and discover the anchor late.

The shower was the second lesson. We test-fit the pan before the plumbing went in. Dry fits catch the mistakes; the real install is where it counts. In software terms, that's the spike, the prototype, the risky integration done first on a throwaway. The teams that test the hard seam early are the ones that don't find it in production.

Third lesson: rework is a phase, not a failure. Wiring day was the biggest single system on the list. I labeled everything, confident and neat. Then I spent the day tracing wires because the labels lied. It was tedious and frustrating and completely normal. Planned holes beat surprise holes. The teams that budget for rework are the ones that stay calm when it shows up.

And the one I keep coming back to: we took the maiden voyage on August 11, with the van nowhere near finished. The list was still long. But the systems that mattered worked, the bed was comfortable, and the road felt like the whole point. We shipped the concept and kept building.

That's the hardest habit for engineering teams to pick up. The demo, the pilot, the vertical slice — shipped before done. Not because it's complete, but because it proves the thesis. Every shop day after that first trip was better, because we were improving something that already worked instead of hoping something would.

A van tour is just walking around your own decisions for four minutes. A program review is the same thing, with a bigger audience. Build the anchor first, test the hard seam early, budget the rework, and ship the proof before you're finished.

The tour isn't the van. It's the decisions.

Think this argument fits your event? Tell me about the room — the calendar is selective.