Fact
You'll never get paid the value you create at a big company
James Hong built a website at HP that eliminated three jobs but only got a $5,000 raise. He realized big companies pay people near the median value any employee provides, not the value an individual actually creates, so the way to capture your own value is to work for yourself.
“And I just remember thinking, like, I'll never make as much money as the value I deliver in a company, because a company, a big company at least, has to kind of pay people based on kind of like the median value that any given person at the company will provide.”
Steal thisIf you consistently create far more value than your salary, stop waiting for a raise and go capture that value working for yourself.
Story
The smart Berkeley kids went to HP; the weak students went to eBay and Yahoo
Out of Berkeley in '95, the top students took 'safe' jobs at Intel and HP while weaker classmates took jobs at unheard-of startups like eBay and Yahoo. When Yahoo IPO'd around '96, those friends got rich, and Hong realized he had been playing the wrong game.
“And meanwhile, our friends who were not as strong students who couldn't get into a good grad school, who couldn't get a job at HP, they ended up taking jobs at other places that we'd never heard of, like eBay and Yahoo or whatever. So like around '96 or '97, maybe '96, whenever Yahoo went public, all these friends got rich, and that's kind of like when you realized that like, oh, like— What game should I be playing?”
Story
Hot or Not launched on a free E-Trade PC hidden under a desk
Hot or Not first ran on a junk PC Hong got free from E-Trade for opening a $500 account. As traffic exploded they scrambled to a Sun machine and leased through Rackspace, with Larry Page personally tipping them off to use cheap Rackable rack-mount systems.
“Actually, when we started Hot or Not, when it first started taking off, it was built on a PC I got for free from E-Trade, I think, for opening an account with $500 in it. It was like the dingiest machine. It had no memory or anything, but it could run as a server. Not very well, but it got us online, right?”
Story
Hot or Not was Tinder in 2000, when posting a photo online was taboo
Hot or Not let people submit a photo and be rated 1-10 by strangers, a decade before Tinder. Hong notes the radical part wasn't the rating but posting a photo at all, since at the time photos lived only behind password-protected Shutterfly pages.
“And this was kind of in the day when everyone was scared to post their photo online. Like if you posted your photo online, it was behind a password-protected page that, you know, was for Shutterfly or Ophoto back in the day. So the concept of posting a photo for other people to see that you didn't have full control over who could see it was completely foreign at the time.”
Story
The Turkish Stud: the inspiration for Hot or Not (and maybe Borat)
A Turkish man's innocent ping-pong photos were turned by an anonymous prankster into a fake 'come for sexy time' page that went viral via StumbleUpon-like site eToro. Hong believes Borat was based on him, and the accidental virality inspired Hong and his co-founder to build Hot or Not.
“He was— he called himself the Turkish stud, and he's like, who wants to come to my country? I can invite you, come have sex. You know, like, it was Borat basically, like, oh, come for sexy time, right? And, uh, and, and it blew up. There was this company called eToro that kind of like, um, jumped on it”
Take
Build something that goes viral, not something that makes money
Hong's founding goal wasn't revenue but virality. Watching the Turkish Stud get on Letterman by accident while well-funded startups couldn't buy attention, he and his co-founder set out specifically to 'build a Turkish Stud' of their own.
“We just thought it was funny. And so we kind of had in the back of our mind like, someday I want to do— I want to build a Turkish stud. Yeah, right. So it wasn't even about money, it was about I want to do something that goes viral, right?”
Steal thisPick a project goal of pure virality first; figure out the business model only after you have attention.
Number
Hot or Not hit 30,000+ unique IPs on day one from 42 emails
Hong seeded the site with himself, his co-founder, and photos from Excite Personals, then emailed a link to 42 friends at 2pm. By end of day the logs showed over 30,000 distinct IP addresses, likely 100,000+ actual people given how many used AOL and dial-up proxies.
$30K
Distinct IP addresses on launch day · unique IPs
“it was 42 people, which I counted it later, but you know, 'cause 42 is a magic number, right? Yeah, of course. But anyway, so I sent it out and like, it just immediately took off. And I think by the end, that was at 2 o'clock in the afternoon, and by the end of the day, we had had like, 30-something thousand IP addresses in our logs, distinct, one unique.”
Fact
Bandwidth is billed at the 95th percentile — spike for free
Hong explains an old bandwidth-pricing hack: carriers bill at the 95th percentile of usage, so 5% of the month (about a day and a half) you can spike traffic to infinity without paying more. Hot or Not exploited this gap as costs threatened to hit $40-50k/month.
“But back in the day, like when you pay for bandwidth, most carriers will give you— will bill you at the 95th percentile of your usage, which means you have 5% of— you can spike to infinity and you're fine, right? So that basically is a day and a half. In 30 days, right?”
Fact
Bandwidth is billed at the 95th percentile — spike for free
Hong explains an old bandwidth-pricing hack: carriers bill at the 95th percentile of usage, so 5% of the month (about a day and a half) you can spike traffic to infinity without paying more. Hot or Not exploited this gap as costs threatened to hit $40-50k/month.
“But back in the day, like when you pay for bandwidth, most carriers will give you— will bill you at the 95th percentile of your usage, which means you have 5% of— you can spike to infinity and you're fine, right? So that basically is a day and a half. In 30 days, right?”
Tactic
Offload your biggest cost onto someone else (let Yahoo host the photos)
Image bandwidth was nearly Hot or Not's entire cost, threatening $50k/month and doubling. Their fix was to stop hosting photos themselves and have users submit URLs to photos hosted on Yahoo/GeoCities, killing almost the entire bill by making Yahoo pay for it.
“—so the first thing we did was, that's when we stopped hosting photos and we started sending people to Yahoo and saying, "Hey, just send us the URL of the photo on Yahoo and let Yahoo basically pay for this," right?”
Steal thisFind your single biggest variable cost and engineer a way to push it onto a free third-party platform.
Fact
Double opt-in dating: no rejection in a liquid marketplace
Hot or Not invented double opt-in dating (the model Tinder later used): a match only happens when both people say yes. Hong argues a highly liquid marketplace removes rejection entirely because you don't even remember everyone you said yes to, and it protected women from being flooded by paid messages on sites like Match.com.
“So we were the first ones to do like the concept of what you would call double opt-in dating. Like if you, if you say you're interested in someone, then they can say they're interested in you too. You know, like what Tinder is today.”
Framework
The laundromat business: any conversion model plus a lean operation wins
Hong's takeaway from Hot or Not is that scale doesn't matter if any model converts even a small percentage of users and you stay lean. They became 'the modern-day equivalent of people who ran a laundromat' — just keeping machines running and collecting quarters.
“And then, you know, it doesn't matter if you have scale, if you can even have any model that makes enough money, if you are a lean operation, you can definitely pay your salaries and have like a lifestyle. Business, if you want, right?”
Steal thisTreat a profitable, low-maintenance product like a laundromat: keep ops lean, collect the quarters, and don't chase scale you don't need.
Framework
The laundromat business: any conversion model plus a lean operation wins
Hong's takeaway from Hot or Not is that scale doesn't matter if any model converts even a small percentage of users and you stay lean. They became 'the modern-day equivalent of people who ran a laundromat' — just keeping machines running and collecting quarters.
“And then, you know, it doesn't matter if you have scale, if you can even have any model that makes enough money, if you are a lean operation, you can definitely pay your salaries and have like a lifestyle. Business, if you want, right?”
Steal thisTreat a profitable, low-maintenance product like a laundromat: keep ops lean, collect the quarters, and don't chase scale you don't need.
Number
Hot or Not did $6M in earnings before they sold it — out of exhaustion
Run for 8 years by just two founders out of a house, Hot or Not was doing roughly $3-4M in revenue (almost all profit) by year 3, and about $6M in earnings by the time they sold right before the 2008 crash. They sold mainly because they were tired of managing it, likely for less than it was worth given the downturn.
$6M
Annual earnings at time of sale · USD/year
“Because at that point in time, it was doing like $6 million in earnings. But both of us, frankly, were so tired of it.”
Take
Don't sell out your youth — time with no liabilities is the real asset
Hong's advice to a 21-year-old: stay frugal, don't borrow to live lavishly, team up with co-founders who'll live cheaply, and avoid raising money even though it's easy. When you're young you have no mortgage or family, only time, and you'll never have time with zero liabilities again — so don't trade it away cheaply.
“Because when you're young, you have no liabilities. Like, you don't have a mortgage. All you have is time. Right. You're drinking 5 days a week. But the reality is time is all any of us have.”
Steal thisWhile you have no liabilities, keep your burn near zero and bootstrap rather than sell your youth or your equity early.