My background is building technology startups. In this world, there’s a term called “lean startup,” where the idea is to build a product and get it into the real world as quickly as possible, then making changes from there.

The philosophy is that one will never know what people want, and the only way to find out is for them to “tell you” directly through their use of the product (or not). It’s a useful heuristic process for building scalable, easily changeable software products. Unfortunately, I’ve learned that this doesn’t work with the world of “brick and mortar.”

The biggest lesson I have learned as a DIY builder is to have a plan—a detailed plan. Although it will inevitably change as you go, it helps to lock down your vision and create a framework before construction begins.

And when I say a plan, I don’t mean just a floor plan: I mean an entire architectural and design plan that flows into your budget. This means you hypothetically know what the place will look like, what materials you will use, where you will get them from, and how much it will cost. Your construction process will inevitably have its own issues, but if you have a plan, you can focus on those new problems and make game-day decisions within the framework of your plan.

There have been a number of design decisions throughout the building of Hawkeye House that I wish I’d thought through a bit more. For example, I got very “concrete happy” in the hardscaping process, and could have mixed in other materials as well—more flagstone, natural dirt, etc. Along with concrete, I think I used the bulldozer a bit more than I should have. Once you bulldoze something to make a flat area or a pad, it becomes difficult to make the surrounding areas look natural again.

Having said all of that, the biggest blunder I made would have to be the budgeting of my electrical system. In the end, it actually works very well, but I spent roughly $3,000 more on it than I’d planned.

In an off-grid solar system, one usually has a backup generator to charge the batteries in case there’s a day without much sun. Having used generators in the past, I knew they make a lot of noise, so I wanted the dome’s to be as far away from the building as possible—about 150 feet away. What I didn’t fully realize is that when you put an AC source relatively far from the main panel, you need very large copper cables to transmit that power to avoid voltage drop.

The difference between 15 feet and 150 feet of cable is about $2,000 in copper. That doesn’t include the 18- to 24-inch trench we needed to dig, the conduit, and the labor required to route the stubborn copper cables through 150 feet of conduit. Not only was it a lot more expensive than I had planned, but it was also a lot harder than expected—trenching and laying cables is not easy business. In the end, this did deaden the sound of the generator, but I definitely didn’t have that in my original budget.

For Dwell+ subscribers: click here to read the full home tour.