A few weeks ago I was hosting an informal chat with some interns at our San Francisco office. I try to do this at our major offices throughout the summer because it's important that every person working on our business (even if it's just for a few weeks) understands how their contribution connects to our mission.

At these gatherings, I often get the same questions--where the idea for Zillow came from, how we make money, advice for being a great intern--so it surprised me when one intern shot his hand up and asked something I'd never heard before: "What are the challenges of being a tech CEO when you don't code?"

I was both impressed (bold for a 20-year-old) and taken aback, but it's a fair question. I don't know how to physically architect our website because I don't know how to write code. The closest I've come was my hardcore financial modeling in Excel (Wall Street's version of coding) early in my career.

So if I don't know how our product is made, how can I steward its development?

The knowledge gap is especially acute when addressing our product backlog. When a tech leader says something will take six months to build, I can't argue with that--even if the timing frustrates me--because I don't actually know how long something should take to build. Is six months quick for a feature of this type? Is it slow?

I have no direct knowledge to base my assessment on other than experience and impatience, and that's a significant vulnerability when my job is to keep the company moving fast and staying ahead of competitors. But my impatience can't win out--it's not rooted in anything concrete.

Where companies run into trouble is when executive mandates trump what is physically possible. Undoubtedly, the Volkswagen leadership who set the impossible specs for its engineers to live up to didn't actually know how to build the car.

Because the architects of the product were asked to make the impossible possible and lacked permission to push back, bad things happened: The engineers cut corners and cheated, compromising their product and brand.

I don't want to create that snowball of bad decision-making simply because I don't know how to code. Solving for my technical knowledge gap means I need to do these five things:

1. Trust my technical leaders.

When I'm told something will take six months to build, I can question the possibility of speeding up the process but I can't demand it. I need to accept the word of my team, trust they're living up to the pace our culture expects and empower them to help their orgs build great products.

2. Get my technical leaders to trust me.

Skills like strategic insight, people leadership, and the ability to galvanize separate teams around a common goal are paramount. I need to demonstrate that I know what I'm doing as a leader, even if I don't know how to build the product.

3. Proactively ask about technology investments.

With a non-technical background, it's easy to get caught up in the familiar or flashy aspects like revenue generation and feature development. Code architecture, improved process and tooling, rewriting an aging code base, and raising our quality and reliability bar are just as important.

4. Focus on the macro technical issues.

This helps me understand where engineering teams are coming from. It also helps me to have more meaningful discussions around the resources needed to clear their roadblocks.

5. Keep learning.

I've been in the tech industry for most of my career, so a certain amount of technical prowess sinks in over time. But things change quickly, and monitoring the pulse of that change--like knowing the latest code languages and platforms so I don't sound like a dinosaur--doesn't require an engineering degree.

Perhaps the fact I don't know how to code is due to a generational gap. I graduated from Harvard in 1997 with a degree in government; I never learned Ruby or Python (though apparently, everyone is updating to React).

Looking back, it's a travesty that I could graduate without taking any coding classes. It's becoming as important a language requirement as Spanish or Mandarin--and perhaps even more so as our translation abilities exponentially improve through artificial intelligence. We may reach a time where it won't matter that we speak different languages so long as we have a common technical language to build from.

But even if we all know how to code someday, we'll still need leaders who are strong in diverse, complementary skills. We'll need a chief technology officer who knows how to shepherd great product. We'll need a chief business officer who knows how to make that product make money. And we'll need a chief executive officer who knows how to knit these different efforts--and many others--together to propel the business forward.