Email  Print 

more capital articles

• Starting Up: Will AmEx Spread the Wealth?
• Sellers Offer Financial Hand to Buyers
• Small Firms Get Local Loans
• SBA-Backed Loans Dry Up
• 3 Sources of Franchise Financing

PALO ALTO, CALIF. — To bootstrap his start-up, HotOrNot.com, James Hong had to go to extreme measures, including what might be politely called borrowing bandwidth from University of California-Berkeley and hosting services from Yahoo Inc.

But that kind of resourceful attitude is what is necessary for an entrepreneur to survive and succeed, according to Hong, who dispensed advice on bootstrapping and revenue models Thursday night at Startup2Startup, an event in Palo Alto, Calif., that is designed to connect beginning entrepreneurs with experienced ones.

Hong's HotOrNot.com, launched in 2000, became an Internet sensation, providing an early example of viral marketing and online advertising revenue before Google AdSense. The company never took any funding and sold earlier this year to Avid Life Media for a rumored $20 million.

Many entrepreneurs bootstrap their start-ups, going without outside investment and using their own resources, such as savings or credit cards, to fund their businesses. The reason is often simple: They can't secure funding from outside sources, such as banks or venture-capital investors. The upside of going it alone is maintaining control of their companies. Other successful bootstrapped technology companies include the free classified Web site Craigslist Inc., enterprise and consumer software firm AdventNet Inc., and PlentyofFish Media Inc., which operates the most popular U.S. online dating site, according to research firm comScore Inc.

On a lark, Hong and his partner Jim Young in October 2000 launched HotOrNot.com, which allows people to post photos of themselves and rate others on a scale of 1 to 10. After sending an email to 40 friends to announce the launch, it had 40,000 unique visitors in one day. Eight days later, it had 2 million page views.

Hong emphasized that it is important for entrepreneurs to spend as little money as possible. He ran the company out of his living room for four years, by which time the company was generating about $4 million per year in revenue, almost all of which was profit.

"The stories I'm going to tell you about are the past," Hong said. "It's not going to be applicable to your specific start-ups because times change. But what I want to convey is how cheap and scrappy we were."

In 2000, bandwidth costs were about 100 times as much as they are today, so Hong says he almost shut down the site early on because he calculated that the site could cost about $100,000 per month to operate.

To address this, when users clicked to upload their pictures to HotOrNot, the site diverted them to Yahoo Inc.'s free Geocities service to host them there.

"I remember having this idea at In-n-Out at 1:30 in the morning: Why don't we just FTP all the pictures to Geocities?" Hong said. "We just ask users to submit pictures to Geocities and we refer to their URL and therefore we don't have to pay for the bandwidth."

Yahoo later shut down other sites that were pilfering Geocities as a storage service, but strangely they didn't shut down HotOrNot. Hong later found out that people at Yahoo really liked HotOrNot and did not want to be known as the company that killed it.

Later, HotOrNot signed up with online photo services Ofoto and Shutterfly, which were seeking more members, to have HotOrNot host some photos on those sites. HotOrNot sent leads to those sites, and was paid for it, thereby turning one of its largest expenses - photo hosting - into a revenue source.

To save on bandwidth, Hong and Young pilfered it from UC Berkeley, where Young was a graduate student. The pair stashed a computer server for the site under a desk and piled some books around it to hide it. Eventually they got caught and had to find bandwidth elsewhere.

Later the pair signed favorable deals with hosting company Rackspace Hosting Inc., convincing the company that HotOrNot would be the perfect test case to show that Rackspace could scale a Web site to large proportions. Rackspace, which was relatively new at that time, was willing to take the risk on HotOrNot. (RackSpace last month set the terms of its pending initial public offering at 12.7 million common shares with an estimated price range of $12 to $16 a share.)

"HotOrNot never paid for a machine or for bandwidth for its first year," Hong said.
It's all about survival for start-ups, Hong said. Which means spending as little as possible, especially in the early stages.

After the site grew, Hong decided to add a subscription service in addition to the free ad-supported model as ad rates dropped from 10 cents per 1,000 page views to 2 cents.

Hong created a dating system where if a person saw a photo of someone he was interested in, he clicked a button. The other person would then have the option to agree. Only then, when there was a "double match," would the site ask the users to pay in order to communicate.

HotOrNot had a conversion rate from free to paying users of more than 15%, which was much higher than other dating sites, Hong said.

"How did we do it? By the time you even knew that we were charging you money to talk to this girl, she already said she was into you. So it was a pretty compelling sell," Hong said.

The pay model was based on the idea of buying drinks at a bar.

"If you wanted to write her, one of you had to be a member," Hong said. "At the end of the day someone has to buy drinks, and usually it's the guy."

HotOrNot was also one of the first sites to sell virtual goods, Hong said. People could buy virtual flowers for each other for a cost of about $2 to $10. "They're not really buying the flower; they're buying attention [and] prestige," he said.

There is no simple rule for when to make a site free versus a subscription. But when a site is not very differentiated or unique, Hong said, it's better to go free, because otherwise someone will try to offer the same service for a cheaper price.

One of the reasons Hong decided to sell was because he felt the site should be free to compete better and therefore be focused on brand advertising, in which he had no experience. That was one area, apparently, where he was not willing to jump in blind.

Write to Tomio Geron at tomio.geron@dowjones.com