In the old days, tech companies faced the substantial capital expense of buying lots of expensive hardware, not to mention hiring IT people to maintain them. Animoto represents a new type of company — one that outsources virtually all of its computing needs. Animoto is a Web service that automatically creates videos from your personal photos and favorite songs. According to the company, the site has drawn 250,000 registered users, more than 10 percent of whom upgrade to paid service, and another two million have used the free Facebook application. And in July, Animoto was the top Facebook app in terms of active percentage of users.
How did you start the company?
Hsiao: Literally it was over heavy drinks and lots of napkins in the back of one of our favorite sake bars in New York. My friend, Stevie Clifton, and I would get together and shoot the shit. I was producing shows for MTV, Comedy Central, and VH1. Stevie was working over at ABC doing documentary work at the Peter Jennings group. We're particularly interested in how we might use technology to automate creative decision making, specifically automating some of the stuff that editors, directors and producers do in the editing room. There's such a gap between the quality of videos on the Web and stuff you see on TV. Most of the stuff you see on YouTube looks like a third grader shot it with a camcorder. We work with the most cutting edge technologies in TV and film, so we know how easy it can be to make stuff look more professional. We really wanted to give everyday people the ability to make more professional videos.
How did you assemble your team?
Hsiao: We worked on this technology during our spare hours, late nights and on weekends. Over the course of several months we realized that what we wanted to do was actually possible, so we quickly put our team together. The company was founded with four guys. Brad Jefferson, now our CEO, brings a whole history of incredible business experience. We wanted to make sure the business was developed from day one along with the product. We did not want to be one of these companies that developed a core technology and three or five years later tried to cram a business model onto the technology. We see so many businesses where the business model is an afterthought. Tom Clifton, our creative director and Stevie's brother, was just wrapping up a masters in music at Cambridge in England. Like Stevie, he's a technical genius. The four of us were the perfect team.
You guys have known each other since high school and three of you went to college together. What's it like working with your teenage buddies?
Jefferson: There's no politics that play into the relationships. We're all able to be very frank and honest with each other. If we don't like something, we're willing to air that right away and get it on the table.
How did you turn the idea into a product?
Jefferson: We founded the company in August '06. We had this idea that we'd be able to get some sort of working prototype finished by the end of the year. It was really important to hit that milestone because our wives and girlfriends were looking for something tangible for what we were doing with our time. We planned this Christmas party and we wanted to have this great prototype to show them and convince them that, yes, this was a good use of our time. At the holiday party we finally got the first video to render. We showed it to ladies and the video played upside down. We're thinking, "Oh my gosh, we hit the jackpot! We got something to work!" Our significant others were rolling their eyes like, "Oh man, they've got a long way to go."
Why did you decide to outsource all of your computing?
Jefferson: Once we launched the private alpha in March '07 to family and friends, the feedback was very good. I started modeling what this would look like in terms of the systems required if it was as popular as we thought it might be. What was very evident was that buying the infrastructure would be very expensive. We made the decision that we would re-implement the entire stack on Amazon Web Services in lieu of just using a regular hosting provider. That was a tough decision. We had to delay our launch by three months in order to do that. Obviously that would open the door to any competitors to arise in those three months.
You saw a big jump in traffic after you launched a Facebook app. What happened?
Jefferson: In the first month it grew to 25,000 users. During that month, we were playing with lots of different viral hooks Facebook offers you for messaging and creating interaction with friends. In April, we stumbled upon the right combination of making this system viral. Within one week's time, we grew from 25,000 users to 700,000 users. In order to support that kind of rapid growth, the number of servers required to create these Animoto videos scaled from 50 to 5,000 in the same period. If we actually had servers in our own serving room or if we were using a hosting provider, it's impossible to have 50 physical servers and grow it to 5,000 servers in that period of time. It's insane. Assuming that a processor costs $400, 5,000 processors would be about $2 million. That would have meant taking venture capital for at least $2 million and we would have lost a lot of equity in the company.
What other services do you outsource?
Jefferson: We use Amazon Web Services for all IT infrastructures like processing, storage, bandwidth and queuing. We use PayPal and Google Checkout for billing/payment and outsource our DVD creation to another vendor. We also use SaaS services for things like e-mail (Google Domains) and Sales Force Automation (Salesforce.com).
So how much infrastructure do you actually have?
Hsiao: The only real asset we have in our office, and the biggest thing we bought recently, was a fancy espresso machine. We have our own personal computers, but no servers and no technical infrastructure.
That lets you focus on product and innovation instead of housekeeping?
Hsiao: Exactly. We're afforded the luxury of focusing on what were actually good at, which is video creation. We don't want to spend a good chunk of our time maintaining a bunch of servers, adding servers, fixing broken ones and so on. Our product is so processor intensive because we're rendering all of these videos from scratch. Quite honestly, a company like ours could not have existed a couple of years ago without cloud computing. The way we use processors is all over the place — usually we're using around 50 and sometimes it jumps to 5,000 or 8,000. There's no possible way we could handle the variances of our processor usage without quickly digging ourselves into a multimillion dollar hole.
So what's your business model?
Jefferson: We explored advertising as an option and it just didn't work. We decided that we would make a freemium model, which allows people to get a taste of the Animoto service for free. You can create a 30-second video for free but if you want to create a longer video it's $3 per video or $30 a year for as many extended length videos as you like. You can upgrade one of your videos to DVD quality for $5. For $20 we'll send you the DVD. In June, we launched another service called Animoto for Business. In May, Amazon invested some money, along with family and friends.
Have you been profitable?
Jefferson: Not yet. We're actually going through the process of figuring out where we want the future to take us, whether we stay the course or bring in a little bit more capital to lever up the business. We feel like we have great competitive advantage right now so there's a desire to move quickly to capitalize on our position in the market.