Skip to main content
Management in Practice

How Do You Build a Startup?

How does an idea become a profitable company? Few people have more experience developing a venture than serial entrepreneur Kevin Ryan YC ’85, founder of Gilt, MongoDB, and Business Insider and former CEO of DoubleClick. Ryan describes his approach to evaluating ideas and assembling a team.

Launching a business is hard. About half survive five years, according to the U.S. Small Business Administration. At the 10-year mark, just a third of firms are still around.

And the odds of hitting it big? CB Insights estimates that 1% of venture-funded startups reached unicorn status—a billion-dollar valuation.

To understand what it takes to reach that point, Yale Insights talked with someone who has done it multiple times. Kevin Ryan has played a significant role in developing New York’s tech scene, building companies and helping to assemble a network of investors and entrepreneurs. He joined DoubleClick in 1996 when it had a dozen employees, and grew the business until it was sold to a private equity firm for more than $1 billion in 2005. He has since founded companies across a range of industries, telling SmartCEO magazine that rather than focus on a given sector, he looks for ways that digital approaches can improve existing industries. He is currently founder and chairman of MongoDB, AlleyCorp,, Workframe, and Nomad Health.

Q: How do you decide which opportunities to pursue? Are you looking at whole industries or specific ideas?

In general, my approach is bottom up. I look for a problem to solve. Any time I run across a situation where I’m thinking, wow, this is hard to do or this takes a long time or this is incredibly expensive, that’s a potential business. Solve one of those three problems and you’re off and running.

In terms of big areas, I’m intrigued by everything in education, healthcare, and fintech because they have not been completely transformed, yet. With areas like retail, a lot of things have been done, so I see fewer opportunities.

Q: You must have more ideas than you can act on. What’s your process for choosing which to move ahead with?

I’m always thinking. I have 15 different ideas that I’m considering right now. There’s no scientific formula. One thing I consider is, am I excited about it? By excited I mean non-financially. Do I want to spend time on this idea for five or sometimes ten years? I have to be intellectually interested in it.

Also, do I think it can be a big enough business that it’s worth spending time on? Some things are good ideas but they are just features or products, not really companies. Most importantly, I must have a clear vision of the product. I don’t care about financials. I need to know what my product is going to look like in my mind. If that’s distinctive, that’s what makes me feel ready to go ahead and launch.

Even then, it’s inherently risky. You’re going to have to make a leap of faith because if it were obvious, other people would’ve already done it.

Q: Does market research play a role?

I don’t do market research in a really structured way. I do talk with potential customers or clients a lot so that I understand their thought processes. But many times, I’m offering a product that is new enough that people won’t know whether they need it until they can use it. It’s almost like a novel. I can tell you it’s a boy-meets-girl story, and you say, “I don’t know if that’s going to be good.” If you can read the book, you’d say it’s fantastic, but the idea itself isn’t compelling enough; it’s how you do it.

Q: As a serial entrepreneur, how do you figure out the right level of involvement in a company?

I was involved with my first company, DoubleClick, for nine years. I was the CEO for most of that time. As CEO you can’t have many other commitments. Once we sold DoubleClick, I could start more than one company. I had ideas and I thought it’d be incredibly fun and intellectually interesting to follow up on a number of them. That was also a way to diversify my risk.

One dangerous thing about being an entrepreneur is spending four years on an idea that ultimately doesn’t work. That’s very costly, but I was highly confident that if I launched three things over three years at least one of them would work. I’ve kept using that approach, and it has turned out that almost all of them have worked, and so it’s been really fun.

Q: How do you know when it’s time to add something to your portfolio without spreading yourself too thin?

I have six companies right now that I’m founder and chairman of. Added to that, I’m an early angel investor for about another five companies. I’m quite involved with them but not as involved as the ones that I founded. It’s working fine. I allow for some buffer time in case I need to focus on one company if something is going wrong. For example, if I need to change CEOs. That buffer, which is probably 20% of my time, goes to working on new ideas, meeting new companies, or tangential thinking, where I don’t know where it’s going to lead but once in a while results in an idea that’s really good and I feel compelled enough to start something.

I get an idea for a new company about once a year, but there’s no schedule for it. I started Business Insider and Gilt literally the same month. I just had both ideas at the same time. After thinking about it, I realized that I had a crush on both. Nine years later, I was still the chairman of both and then, a year later, I sold them both.

Q: What do you look for when choosing CEOs?

The relevant skills are very different because my companies are in very different areas and they’re very different sizes. At a high level, any CEO needs to be a leader, but it’s the equivalent of choosing captains for the rugby team and the gymnastics team. It’s not the same person.

I start by saying, “What is the core skill of this company?” I’d like the CEO to have that skill. I have a healthcare company where I wanted the CEO to be a doctor. For Business Insider, I needed someone who could be editor in chief as well as CEO. For my e-commerce companies, I needed someone with e-commerce skills.

In the beginning, the CEO of MongoDB had to be extraordinarily technical, because all we did was deeply technical work. Now, our CEO can be a little bit less technical and more sales-and-marketing oriented because our technology’s very good. So what’s important at that moment plays into the choice.

Q: Are you looking for previous experience leading a company?

If you’re risk averse, you’d say, “I want someone who’s already been a CEO multiple times.” Often, I lean toward first-time CEOs. I want someone who’s super smart but ready to improve themselves because they are feeling a tiny bit nervous—they feel 10% out of their league. That’s where you get the best work out of people.

The things they don’t have are the things I do have. Maybe they haven’t done an option plan before. I’ve done 50, so I can help. What they can do that I can’t is put 60 hours a week into selling, marketing, hiring, and being out there doing all the hard work that builds that company. I want people who are very passionate and feel like they’re going to overcome the odds. Most startups don’t make it. Very few become big. I want all my companies to become big.

Q: Do they need to come from a startup background?

It’d be really rare, in fact almost never, that I would hire someone who had been at IBM or any big corporation for 15 or 20 years. They’re not going to have the right approach in general. I would rather someone who has been working for a startup that’s grown from zero to 300 employees. They may be the CTO or the head of business development or sales, and they’re ready to take that next step. They have already established a record that they are good at managing a small team, and now they can make the next big leap.

Q: Which is more important, the idea or the execution?

Execution. Obviously, the idea is important but not nearly as important. Ideas evolve, which is really part of execution. You start going down one path; then you realize that you need to make a change. Sometimes it’s a small alteration. Sometimes it’s big enough you’d call it a pivot. At the end of the day, what’s important is hiring great people and making good objective decisions along the way. Over time, the team will adjust the idea whereas the idea won’t adjust the team.

Q: Looking across your companies, is there a consistent culture?

The culture’s going to be driven largely by the sub-industry. When you get 200 journalists in one room, that is inevitably a different culture than 200 engineers who are working on a database. Beyond that, the culture reflects the leadership. It reflects me a little but more the CEO. We consciously try to create something that’s fun, dynamic, interesting, and transparent.

I think a lot of it is how you’re treating people. For every company you’ve worked for, you have an impression of the senior people leading it. You had a sense of whether you like them, whether they’re smart, dedicated, focused, transparent, honest, effective. That’s what creates the environment and culture.

We’ve all worked for someone who was very good on the technical substance of their job, but they were not nice people and not fun to work with. Working with those type of people makes you hate your job over time. You need to get rid of them. I think some organizations are not tough enough about solving for culture versus technical competency. But I don’t have a lot of concerns about the cultures that we’ve created.