In the most recent episode of Startup School Radio, host and YC partner Aaron Harris sat down with Elad Gil.

Gil is an investor, entrepreneur, and veteran of Google and Twitter who is now the founder of Color Genomics, the startup that’s building a platform to deliver affordable genetic testing to screen for the risk of diseases such as breast cancer.

You can listen to the entire episode in Soundcloud right here, or in iTunes here.

One particularly interesting insight was hearing Elad talk about how hiring is key, particularly when your company is tackling complicated problems:

Aaron : The level of complexity [for a biotech startup] is so much higher than a software startup or even a normal hardware startup…. How do you go about building a lab as a startup? How do you even know where or how to start?**

Elad : We basically hired experts. We hired people who’d built labs at the Penn Genome Center, and had been working in CLIA certified labs at Illumina and other leading companies. Really, the emphasis was, “How do you hire people who have done it before, who understand the regulatory environment, who understand quality controls, who understand how to do these things very very well?”

There’s often startup advice given that I don’t 100% agree with: “Just hire really smart people and just have them do whatever.” I often think that’s often a very bad strategy.

I think what you want to do is figure out, where do you need very deep expertise? And then, where do you need people who are just adaptable and smart? …We always talk about that. Our company asks, “What’s that person’s superpower? What’s that thing that they’re uniquely great at?” And then, on top of that they need to be adaptable and smart and a good culture fit and great to work with and collaborative and all these other things.

But we need to make sure that the one thing that they are part of, they do very, very well. Because that makes a huge difference, so that we don’t have these steep re-learning curves where we have to reinvent things. Because there are certain areas where you certainly never want to reinvent something.

The canonical example is, if you’re with a tech company and you want to build a data center. You don’t just hire a smart person to go and build the data center from scratch. They need to actually understand supply chains and power utilization and cooling systems and all these other things.

Startup School Radio is a podcast that features stories and practical advice about starting, funding, and scaling companies.