I work at a company that provides enterprise services, so when I saw the list of YC startup lectures, I jumped straight to Lecture 12, hoping Aaron Levie could answer many of my questions.

Aaron Levie talks fast and is very funny. He opened with upbeat music and got everyone clapping along, then joked that this was probably the high point of the lecture and it was all downhill from there. The idea for Box came in 2004, when Aaron was writing a paper and found it hard to share files. So he built an app for easy online storage and sharing (another product born from its founder's own need). He teamed up with a high school friend at Duke University and started the company. Starting a company isn't as easy as it sounds; without money, you can't even begin. Amazingly, their seed money was $20,000 won from online gambling. After that, they were lucky enough to meet a remarkable investor, Mark Cuban, who invested $350,000. In 2005, that was a sizable angel investment, which shows Mark Cuban's foresight.

What Drove Box's Success

Aaron Levie believes their success came from several major technology shifts. Storage prices kept falling, which lowered the hardware cost of offering online storage and made large amounts of storage affordable for ordinary people. Cheaper hardware alone wasn't enough, though; faster networks and more powerful browsers were the key. Together they determined the user experience of sharing files online. A painfully slow network and a laggy browser can ruin the experience of even the greatest product. With the hardware and software in place, and people all over the world hungry for online storage and sharing, the huge demand ensured the market was big enough. "Any market where something fundamental changes, whether it's the raw materials or another driving factor, is a sign of massive market change." Aaron was lucky enough to find one.

Box didn't target the enterprise market from the start. In 2006, Box was free to use and gained lots of users. It had built a full consumer product, and very few enterprise customers. Aaron faced two paths. One was to keep going after consumers. That was fun, but it was hard to make money: you had to work very hard to get users to pay a few dollars a month, and the ways to monetize were app purchases or ads. The whole market was only worth a little over $100 billion. The other path was the enterprise market, which was huge and stable, worth as much as $3.7 trillion a year, and enterprises cared deeply about efficiency gains from technology.

The market is huge, but the enterprise market also has many problems: high technical complexity, long development cycles, long sales cycles, and a terrible user experience in traditional enterprise software (one app might have 47 buttons). If Box followed the path of old-school enterprise software, it might not get a share of the market, no matter how big it was. Aaron kept thinking about how to make enterprise software different from what people were used to. True to his usual approach, he not only thought about how to change and stand out, but also watched for shifts in the ever-changing world that could help them change, so they could ride those trends and build a better enterprise software company.

So What Are These Changes?

Enterprises and enterprise software are changing dramatically. If you want to start an enterprise software company, focus on technology and solutions. This is the best time ever for enterprise software. Lots of applications are moving to the cloud. Software that only large companies could once afford has become a standardized service every company can use. Costs have dropped sharply, and development and deployment cycles have shortened dramatically. Custom development has given way to cloud subscriptions with no regional limits, so companies everywhere can use them on demand. The business model has also shifted from IT-driven to user-driven. On top of that, there are 1.75 billion smartphones and nearly 3 billion people online. Almost every industry is changing, and enterprises are most likely to take off when two kinds of change happen. First, the raw materials change: costs fall, things become more centralized, and people can use them on demand. Second, customer needs change: omnichannel retail; personalized healthcare, shifting from paying for checkups and treatment to paying to stay healthy; telemedicine and electronic medical records; and global production, editing, and distribution in media. All of these need software and technology to support them. These changes mean personalized customer needs and huge markets.

How Do You Start an Enterprise Software Company?

Experience is built on replaying what happened in the past, and you can't apply it to everything you'll face in the future. But the fundamentals stay the same.

Have a keen sense for disruptive technology. This is the most basic requirement for a tech company, because only technology can create the gap between a red ocean and a blue ocean. Much of what we're doing now was predicted 5 or 10 years ago; lower costs turned once-impossible ideas into reality. Start small, with a small market or a small application, and expand from there. Take ZenPayroll as an example: it started with payroll, moved into compensation management, and gradually expanded into more services. Look for market asymmetries: things big companies don't want to do. Examine big companies' cost structures and find the costs they can't cut. That's their business model, or a way to make money from their customers. Paul Graham, author of Hackers & Painters, wrote that you should live in the future and build what's missing. That's how you break out of existing ways of thinking and keep up with the times. The same goes for work: if there are customers working on the future, the best approach is to work alongside them and explore areas the future needs but nobody has filled yet. Listen to users, but don't do exactly what they say. Build what they need, and filter out the product suggestions that are actually useful. Build a platform, not custom solutions. Always put user experience first; products that are easier to start using have a better chance. Build consumer DNA into the product during development so it can spread on its own. Finally, read three books: Crossing the Chasm, The Innovator's Dilemma, and Behind the Cloud.