Subscribe to San Francisco Magazine

Mod Lux Feeds

Now Playing

On the wings of angels

Somehow, in the middle of the great recession, the Bay Area technosphere is booming again. Diana Kapp explores the spirited new economy and the next-generation investors who are rebooting Silicon Valley ingenuity—and, quite possibly, the American way of life.

The lights dim, and the audience—every angel investor and VC partner of note in the Bay Area (i.e., the universe)—falls silent. For a second or two, it’s almost possible to believe that no one is tapping, clicking, texting, tweeting; everyone is just watching, listening, waiting to be wowed.

It’s Demo Day, the graduation party in Mountain View for the latest crop of entrepreneurs launching from Y Combinator, the hottest startup incubator in tech. Someone once called the event “American Idol meets Wired,” and that feels about right, except that instead of a record deal and a People cover, the stakes could be the next world-shaking Facebook, Twitter, or Apple. “I’m standing in a room with the future of the tech industry,” says 21-year-old hacker Andrew Sugaya, who wears a burgundy MIT track jacket and is here to pitch his company, his eyes as big as those ancient floppy disks that were out of date before he was born. “I don’t even know what to say.”

Despite the recession almost everywhere else, the people in this room are in the throes of a gigantic if still largely undetected (outside the technosphere) boom. And Paul Graham, the 46-year-old painter turned technologist turned geek whisperer who launched Y Combinator in 2005, has made Demo Day a phenomenon, the most thrilling thing going in a Silicon Valley that’s been seriously feeling the recession. Soon to take the stage, one after another, for two-and-a-half-minute pitches, are the incubator’s current group of entrepreneurs: the overachieving technologists behind 36 startups, including the winner of the “most ruthlessly efficient killer” programmer award at MIT, two Rhodes scholars, and a 20-year-old woman who launched her first company at age 13. Graham and his three partners picked them and their ideas from many thousands of applicants (a slot is harder to snag than a place at Harvard); provided seed funding of $15,000 to $20,000 per company to cover ramen, Red Bull, and a place to crash in the Bay Area; then gave them three months to see what they could do—with constant guidance, of course.

What they have done is remarkable, especially if your only point of reference is the instantly bloated startups of the dot-com era. In just 90 days of boot-camp-intensity work, they’ve finalized their products, incorporated into companies, gotten advice from Silicon Valley luminaries, and even gone beta or fully live. A quarter of the 36 companies are already profitable. While this isn’t necessarily saying much—with minimal expenses, all you need to become “profitable” is a small customer base—these outfits are certainly less financially precarious than the 1990s startups headed by Stanford MBA types who burned through millions in cash raised from Sand Hill Road’s storied venture capital firms while having no clue how they’d get revenue. A new model of creating companies, one rooted in per­petual experimentation based on customers’ real-time feedback and likely resulting in more ka-ching and less roadkill, explains why Graham and his partners are already among Silicon Valley’s leading celebrities and why Demo Day is packed. It also explains why the old guard of venture capital—the VCs who helped create some of the most transformational companies on earth—want in on the action. The next gangly, irreverent geek who will be worth billions of dollars and a full-length feature film by the time he’s 25 may be up onstage.

The truth is, it’s now dirt cheap to start an Internet company. All you need is a good idea, some basic coding know-how, and a little money. And guess who has enough money to fund these kinds of ventures in their earliest stages—to be guardian angels of the newest new thing? Well, pretty much everyone in Silicon Valley who’s had even a modest hit in the past 10 years.

So here they all are at Demo Day, looking to place bets. Retired Googlers and PayPalers, of course. Nearly all the top “super angels,” such as Naval Ravikant, who controls millions he raises from other Valley moguls, some of them prolific angels themselves. VC god Michael Moritz, of Sequoia Capital. White-haired Ron Conway, ”the Zelig of the startup world,” as one investor calls him, leans against the side wall, holding court about the 500 startups he’s backed. A bunch of celebrated VC firms—those that put early money into Microsoft, Oracle, PayPal, and eBay—have sent younger associates, as has Marc Andreessen, the Netscape cofounder whose nearly $1 billion in funding, at least some of it devoted to seed investments, makes his firm the center of this universe. Reid Hoffman couldn’t be here, but he already met all the entrepreneurs over dinner. The LinkedIn cofounder put early angel money into Facebook and social-gaming phenom Zynga, which suddenly has 1,200 employees in Potrero Hill (a few predict it could grow bigger than Facebook).

Some of Twitter’s early angels are in the crowd, the seers who were crazy-smart enough in 2007 to see that 140 characters can do a lot more than tell friends what we had for lunch, such as broadcast accounts of a hotel bombing in Mumbai seconds after it happened. They are sitting on riches (Twitter is valued at up to $5 billion) but still haven’t gotten a penny out of the investment, although that bet alone has earned them renown—in these parts, as much as or more than the guy sitting in the sixth row. He invested in Flipboard and is now promoting the hell out of another investment, the Twitter-like photo site DailyBooth: Ashton Kutcher. With him is his wife, Demi Moore.

Up front, Graham checks the sound. He didn’t dress for the occasion. He wears his usual khaki shorts, blue polo shirt, and Tevas, a fitting uniform for a guy who no longer has anything to prove. Y Combinator now runs its Demo Day in triplicate to meet investor demand.

Silicon Valley is no longer that Silicon Valley, Graham contends—not the place where VC firms rule alone, not the place where middle-aged Republicans were the only angels, and they would invite their entrepreneur pals to their homes in Atherton or Woodside and write a check. “It’s a venture world where nothing changed for 30 years,” says Graham. Now, suddenly, the angels are barely older than the kids they are desperate to fund, and they meet their CEOs over wine at District, in SoMa, or for a beer at Antonio’s Nut House, in Palo Alto (assuming the entrepreneur is old enough to drink). Goofy idealism fills the air, as does the sense that what’s cool is a computer jock with a juicy idea who sleeps on an air mattress—and that clout comes from being tight with him.

The presentations are rat-a-tat fast. I watch as Andrew Sugaya touts his platform, Teevox, which turns an iPhone into a remote for selecting online movies and shows. The three entrepreneurs behind Rapportive—a free browser plug-in that makes your Gmail account look and act like Facebook—report that they already have 40,000 users. Before arriving, I’d heard so much about the intensity on this island of boom that I’m looking for the now familiar signs of a cynical bubble. While “here we go again” does run through my brain, I’m surprisingly compelled by the scene and the smart people dominating it. Maybe it’s not great that our best and brightest’s big dreams are to create a newfangled online flight-reservation system or a remote designed to make us even bigger couch potatoes. Still,
I haven’t felt this much can-do energy outside of AT&T Park in a long time. All these adorable geeks need is to be believed in, and to be handed $50,000 checks by guys who could be their big brothers.  One Valley veteran told me that this is how venture capital used to work, one entrepreneur helping another—a celebration of pure innovation.

It’s only later that I learn that, despite the anticipation filling the Demo Day hall, nearly half the presenters arrived needing even less than I’d thought. They had already gotten most of the checks they needed, because the coolest of the cool angels had cherry-picked the best deals on the downlow, with Graham’s help. Oh well. Worse things have happened than investors’ cutting in line to make small bets on companies with early traction. Worse things we all remember.

Ron Conway says there is a misconception that “every 10 years, we get a Google. That’s not true.” After Facebook, it took only two years until Twitter, he says, then another year after that before Zynga. “Great companies are being created at a much greater rate,” he adds. In the Bay Area, the homepages of hundreds of small companies boasting iPhone apps, social-media platforms, and mobile-phone enhancers can’t stop shouting, “We’re hiring!”

Tech as a whole has had serious problems since the dot-com implosion, but this ecosystem of lean-and-mean startups has rekindled hope. “Let me contend that this is a historic moment,” Stanford School of Engineering consulting professor Steve Blank tells me. “The business and economic consequences of the Internet are finally starting to be seen. The chickens have just come home to roost.” Then, clearing his throat: “This is the last outpost of Amer­ican ingenuity.” Angel Naval Ravikant, who sits toward the back at Demo Day so he can multitask, agrees that we’re witnessing the reinvention of Silicon Valley. “It is happening with a tidal wave,” he says. “It’s happening right now.”

A handsome man of 36 with wavy jet-black hair brushing his shoulders, Ravikant becomes my guide to this new economy. He is a prominent super angel—a VC firm of one—who, all told, has invested in almost 50 startups, including a dozen Y Combinator companies. (He’s a veteran of seven Demo Days.) Early this year, he sealed his reputation by launching AngelList, essentially Match­.com for entrepreneurs and investors to meet and do deals. The site has attracted 450 angel participants and has led to nearly 100 investments in startups. “I feel like now, because of AngelList, I get into any deal, because I’ve sent out so many deals,” Ravikant says, sounding relieved. He’s a likable guy, very down-to-earth despite his determination to get to the top of the angel pecking order.

Ravikant’s life is a blur of blogs and tweets as he sprints around the conference circuit. He works three devices simultaneously: iPhone 4, iPad 3G, 11-inch MacBook Air. “I work all the time,” he says. “I’m home, I’m driving, I’m at a party—I’m always on my iPhone. If I’m there, it’s my workspace.” That insane pace is necessary because you can’t be perceived as an insider if you can’t keep pace with 23-year-olds mainlining Rockstar. Just last year, because he hadn’t developed early ties with the entrepreneurs involved, Ravikant was shut out as a seed funder of both DailyBooth, the Y Combinator company in which Kutcher invested, and Foursquare, the hyped social-media site where users constantly update their location via smartphone. “[Being cut out] was a huge wake-up call,” he says. “It’s not good enough to make BusinessWeek’s top 25 angels” (which he did). “You need to crack the top 10.” As a rule, angels are unapologetic self-promoters—no group of people has ever returned my calls and emails more quickly.

The term angel implies “benevolent investor,” Ravikant explains. We are sitting in his airy SoMa office space, not far from AT&T Park, smack in the center of web-startup activity. (The locus of startup mania has shifted again from the South Bay to San Fran­cisco, where Twitter, Zynga, and Foursquare’s new West Coast outpost are located.) The concept of benevolence resonates with Ravikant, given his own rather scarring experience with old-fashioned venture capital.

At 25, he was a gifted young technologist who had started the consumer-review site Epinions, which was backed by VCs, the biggest two being Benchmark Capital and August Capital. He can’t tell me much about the Epinions deal because he signed a non­disclosure agreement. But a typical VC deal involves two or three venture firms putting in millions in exchange for more than 50 percent of a company—or, if they get slightly less, demanding “protective covenants” that give them veto power over a company’s ability to raise more money, sell itself, or change management. VCs also take seats on a startup’s board with the logic that young founders are naive and need experienced grown-ups to guide them through the backbreaking work of building a durable, profitable enterprise once the sex appeal has worn off. 

In Epinions’ case, the company eventually became Shopping.com, and according to court documents, the two VC firms prof­ited immensely from its $250 million IPO. However, all but one of Epinions’ founders made nothing because the VC-backed merger rendered their shares worthless. “Power corrupts,” Ravikant says, shoving his hands into his black jeans. “Eventually, when they have too much power, VCs will corrupt your company in subtle ways.” Rather than start another company, Ravikant decided he wanted to become an angel. Light-touch angels have been floating around Silicon Valley for a long while. In the 1960s, businessmen Tom Perkins and Don Valentine started funding their friends and interests as a side hobby; they went on to found venerable VC firms Kleiner Perkins Caulfield & Byers and Sequoia Capital, respectively. Oracle, Apple, and Genentech all started with the help of small checks from individuals.

Unlike VCs, angels pride themselves on not wielding power. They get in earlier, joining together to invest much smaller amounts ($25,000 to $500,000 each) so the stakes aren’t high enough in any one deal to merit meddling with the company. The concept recalls the 1930s, when wealthy individuals backed Broadway productions and early Hollywood films. Giving the keynote at a confab in Austin, Texas, thrown by an incubator called Capital Factory, Ravikant admits, “I think of this as charity work.” It helps that many of the new angels are rich enough not to sweat losses. Having settled a lawsuit against Epinions’ VCs, Ravikant never has to work again.

He started with a dozen personal investments, putting $10,000 to $15,000 (and eventually as much as $100,000) into a few startups. Luckily, Twitter was on that list. In short order, Marc Andreessen and third-generation Silicon Valley VC Tim Draper gave Ravikant funds to invest, while well-known VC firm Sutter Hill Ventures hand­­ed over $10 million, looking for its entrée into Twitter-like scores. Ravikant’s emergence as a super angel with his own $22 million fund made him a pioneer, but in the past year, that path has become a cliché. A glut of ridiculously loaded ex-Googlers and ex-PayPalers, plus techies who’ve cashed in on YouTube, Delicious, and the 40 companies that Google alone has acquired this year, have joined the parade. As young as 30, often “retired” from their first huge successes, they are becoming angels partly to address their very rational fear of losing touch with such a fast-changing industry.

A lot of what they do involves realizing the potential of an idea before others see it, which makes this a game with a slim playbook: Filter out the thousands of bad ideas, follow your gut on the best ones, then act early, spread your risk, and expect to fail. It’s not unheard of for angels to commit money at a first meeting. Super angel Dave McClure, a profane ex-PayPaler whose firm, 500 Startups, is raising $30 million and has plans to start its own incubator, admits to twice having invested without meeting the founders in person, based solely on instinct and references. “I don’t need all the information at the first check,” he says. McClure views much of seed funding as an inexpensive way to see what might develop. One VC labels the strategy “cheaply purchased upside optionality.”

One key to success is whether you are friends with other smart people with well-developed instincts. Entrepreneur Kevin Rose created (but no longer runs) Digg, the news site where users vote on what’s interesting online. Along the way, he has built an enviable portfolio of his own investments, including Zynga, Twitter, and Foursquare. He’s friends with many companies’ founders and says that he and they learn constantly from each other. “It serves me; it serves them,” he explains. That reminds me of the best line about the angel-entrepreneur relationship that I’ve heard, thrown out by an audience member at a conference session on angel investing: “It’s like fellatio—two teenagers just fooling around.”

“Lots of times, these deals are oversubscribed and you’re fighting to get in,” says Rose, who readily admits that he’s not above groveling. “I definitely was hitting up Jack Dorsey to get into Square,” he adds, referring to the Twitter cofounder’s new company, which enables merchants to use their iPhones as cash registers. “I pinged him over email, pinged him again. He was kind enough to carve out a little space for me.”

After investing, Rose visited Dorsey at Square and found that they had plenty to discuss. “We talked about some of the pitfalls of stuff happening at Digg,” he says. “They showed me their new app, and we talked about my business.” If Dorsey or Rose ever need anything—for instance, future funding—they can reach out to each other for help. Rose got into his pal Joshua Schachter’s newest company even after Schachter, the founder of Delicious, refused to tell him the nature of the business. Rose said he didn’t care; he’d still invest. So Schachter reallocated slices to allow Rose and other equally faithful investors to join.

Such incestuousness gets a bad rap, Ravikant says: “Despite whatever we might want, this is still mostly a people business.” He mentions two hot companies he wanted to invest in and says he got the nod only because of his past close relationships with the entrepreneurs—Ravikant had tried to recruit one as an executive and had advised the other on business strategy. Because of those relationships, Ravikant’s two friends gave him a break on their companies’ valuations, so he’d have a bigger potential upside from his share. He returned the favor, he says, by trusting them. When the closing documents arrived from one of the entrepreneurs, he signed without really reading them.

At least compared to VCs, angels are just as lax about expected rates of return. VC firms have a guaranteed way to make serious money: 2 percent management tariffs that earn a firm $6 million simply for raising a $300 million fund, no matter what they do with it. Super angels, on the other hand, make much less on fees, so they have to rely almost entirely on the “carry.” That’s a percentage, typically 20 to 30 percent, that they take out of profits left over after the fund amount is returned.

One of the rare studies on angels’ financial performance, by the Kauffman Foundation, reports that 39 percent of angel investors lost money in the stretch from 1990 to 2007. (Another study, by the University of New Hampshire Center for Venture Research, estimates 25 to 30 percent annual return rates, but it counted only those angels who were profitable.) When I ask Ravikant if he knew that Twitter was a home run when he put down his money, he says, “I did—that’s the good thing,” before adding, “The bad thing is, I think that about every one of my investments.” Angel investing is, he joked in Austin, “probably less lucrative than setting fire to a giant pile of money.” Angels refer to “throwing darts” and “Vegas money.”

When I press Ravikant, he guesses that three-fifths of today’s seed investments in the consumer-tech space will lose money, one-fifth will break even, and one-fifth will make money. But getting this ordinarily straight shooter to talk about return rates is like trying to grab a wet bar of soap. “It’s very secretive,” he admits. “You can’t reduce this to math. If someone is sitting around doing valuation spreadsheets, that would say to me that they don’t get this at all. It’s an art, not a science.”

“Every year, there are thousands of deals,” he continues, “but just 10 that matter. If you aren’t in those, you might as well go home.” Not that a bunch of $50 million “exits”—the term for when a company is acquired or goes public, allowing an investor to get his money out—can’t move the needle. But the only sure way to succeed is with one of the rare superstar companies. Google reportedly returned its seed funders 700 times what they put in. The cash Ravikant put into Twitter in 2007 is now worth $5 million to $20 million.

When I do the math on Ravikant’s $22 million fund, I figure he probably needs exits that bring in at least $47 million for him to personally make $5 million. That could happen, but none of his portfolio companies’ three significant exits so far—a sale to Twitter for company stock and sales to Google for $25 million and $70 million—have carried enormous payouts. Let’s say his fund’s stakes in those three startups was $100,000 apiece at a time when each was valued at the fairly common figure of $5 million, and the shares hadn’t been diluted since; the three exits would have returned the fund just $2 million.

Strangely, Ravikant argues that one key to making money is to be a lemming and to flock to the same startups where everyone else has clustered. The level of “social proof” that comes when sought-after angels like Ron Conway or Dave McClure join a deal helps create buzz and momentum, lifting the odds that a company will be able to raise more money later and take off. As in junior high, being part of the cool group becomes a self-fulfilling prophecy.

In the old days, you couldn’t go anywhere as an entrepreneur if you didn’t have VC backing—companies were too expensive and needed a lead VC’s mark of legitimacy. Today’s most talked-about entrepreneurs can wait to take on VCs, and in the meantime, they can pick and choose the angels they want.

Take Rapportive, one of the hottest companies in Paul Graham’s summer crop. The service mines a contact’s tweets and online biographical information and photos and displays them for you in a Facebook-like format alongside any Gmail message you get from that person. (One blogger called it “big brother’s little helper.”) It was hacked together by 27-year-old Cambridge University computer-science dropout Rahul Vohra, who’s been coding since the age of nine and built the product in six weeks at a friend’s office in England. By Demo Day, Rapportive’s customer base was growing 15 percent a week.

As Vohra and his two cofounders pitched to the crowd, however, they’d already quietly raised $1.2 million from the exact people they wanted, including more than 10 angels and three VC firms with young partners who do early-stage deals. “For each, there is a particular reason,” says Vohra, during a call from London. (He’d wanted to Skype but demurred when I admitted I was a Skype virgin.) They’d sought Paul Buchheit, since he created Gmail. Luckily, Buchheit is a close friend of Graham’s and a Y Combinator investor, so Graham made an easy introduction. McClure was chosen after he touted the company vigorously via tweets without having been solicited, then introduced the cofounders around. As for Ravikant, says Vohra, “basically, everyone we asked just said he’s an awesome guy.” Way back in February, Ravikant had already downloaded the product and had heard about the team from two investor friends. He met Vohra and one of the other cofounders for coffee at the Philz on Fourth and Berry Streets in San Francisco, Ravikant’s morning hangout; this was followed by another meeting and more emailing about plans for the product and Rapportive’s competitors. Ravikant wanted in.

Rapportive demonstrates why the traditional VC seed investment of many millions of dollars often doesn’t work for the new brand of startup, which doesn’t need a factory or an HR department or a SoMa warehouse with pool tables and bunkbeds. “The VC model took off in 1985 to fund semiconductor companies building huge wafer fabs,” says Graham. But now it’s so cheap to get a web business going that companies can show potential investors their products before they’ve raised a penny—and in some cases, they can even show real traffic data and paying customers. Rapportive spent less than $5,000 to launch.

Prototypes are so quick and cheap to build (and rebuild) that the companies can easily “pivot”—the term du jour. They iterate constantly, using feedback from real users. Nearly everything they need to launch can be outsourced, automated, or done in the cloud. In a blog he writes called Startup Boy, Ravikant describes the new ease of operations: “They coordinate with Skype and GTalk and wikis. The company itself is snapped together with outsourced HR, cookie-cutter incorporation, and outsourced finance/payroll. Marketing is done virally, or through SEO. PR is handled through tweets and blogging. Payments come via PayPal.” Moreover, he continues, “what used to cost $1M–2M to set up now costs $10K. What used to cost $20M to go to market now costs $1M.” By growing faster on less money, founders can give up a smaller percentage of their companies to get the necessary startup cash.

This new model has given traditional Valley VCs fits. To be fair, the primo firms are still managing to post profits, but those are slim by historical standards. Before the ’90s dot-com bubble formed, VC firms were earning 30 to 50 percent profits; during the peak, they were doubling their money. Chasing the mega-returns of those years, big-eyed hedge funds began pouring vast amounts of capital into Sand Hill Road, ballooning the size of old-line firms. Then new shops created by B-school and finance types with no experience operating small companies came on like the 17-year locust. “It became suits taking money from suits and giving it to suits,” says Graham. Now, says Len Baker, a partner at Sutter Hill Ventures, “I can’t think of anyone in the top 10 firms who made more than 20 percent a year over the last 10 years—no, make that 15 percent.” The years since 2000 have been termed “the lost decade” because investors in VC funds have taken—that’s right—an overall loss.

The bigger the VC funds grew—$500 million is still common—the less practical or even pos­sible early-stage bets became. The outcomes were too uncertain and the amounts too small to impact the bulging portfolios. So VCs started waiting for a tech company to succeed before committing, at which point it would be valued much higher—thereby upping the price of any stake. A recent example: Even when Benchmark Capital wisely invested in personal-finance software company Mint, it waited until the much safer later rounds. Eventually, Intuit paid $170 million for the company, and Benchmark quadrupled its money. But angels such as McClure, who took on the early risk, received returns of 10 to 17 times their original contribution.

As new Facebooks or Twitters have loomed, a few nimble firms, especially Sequoia Capital, have become regular seed funders. Last year, Greylock Partners hired angel Reid Hoffman and provided him with a $20 million sandbox in which to do the early deals he does best. But other firms have taken angel-like steps that look suspicious to angels. Kleiner Perkins’ John Doerr, who recently raised $250 million to fund social-media startups, announced his plans in October with three billionaires by his side: Facebook’s Mark Zuckerberg, Amazon’s Jeff Bezos, and Zynga’s Mark Pincus. The first investment
made by the fund (in a startup called CafeBots, cofounded by a 54-year-old Stanford professor) was a whopping $5 million—not exactly ramen money. Ravikant believes that such efforts may be doomed, because Foursquare-monitoring, FarmVille-playing technophiles dislike the relatively analytical, cloistered, expensive ways in which VCs tend to operate. “Today’s entrepreneurs don’t want to show up in [their investors’] gilded and mahogany offices on Sand Hill Road,” he says.

To wit, entrepreneurs used to brag about the size of their funding round; now they boast about how little money they need. Rapportive could have raised far more than $1.2 million. But the founders didn’t want to give away any more of the company so cheaply and, like the heads of many new startups, they believe they can shoestring and microfund themselves until the money runs out in a year or two. Only at that point, when the company’s valuation is much higher, will they take on large VC investments if they need them. Ravikant believes that some of these startups will be so successful and inexpensive to operate that they’ll never have to go back to the financing well—meaning they’ll never have VC hooks in them. That would be a very good thing, says Graham: “The reason things are moving this way is because the old way sucked for the founders.”

Graham would know: Like most angels, he is a bona fide geek and a former web genius with a profoundly entrepreneur-centric view of the universe. In 1995, he founded Viaweb, an online-store-building app that he sold to Yahoo! in 1998, pocketing $50 million. But he’s not a typical quant jock, having always gravitated toward creative enterprises while operating with a philosophical mindset. After getting a PhD in computer science at Harvard, Graham went to art school. He believes that software coders and painters have the most in common out of all people, and his website includes dozens of essays he’s written on hacker culture, entrepreneurs, and life. As the angel economy has gone from Facebook fast to 1,000-launches-a-month insane, Graham almost gloats about how well it’s going for the top young entrepreneurs.   

Graham’s vibe may be mellow and his hair unkempt, but he’s like a moth to the limelight, and his incubator has become a cult of person­ality. His chosen entrepreneurs all call him PG and drop everything when he calls. He has an idea every second and an analogy for every situation (on the negative effect of angel investments on VCs: “It’s like eating snacks spoils your dinner”). Sometimes he is impressed with a team of founders but not with their idea, at which point he will spew forth a litany of intriguing business suggestions. Andrew Sugaya’s iPhone remote was thrown together at first to ease navigation on his now half-forgotten initial concept, but Graham encouraged Sugaya’s team to seize on it as the capital idea. The original concept matters “almost not at all,” Graham insists. He uses it mostly to understand how an entrepreneur thinks.

Over time, Graham has developed a set of indicators to help him pick his entrepreneurs—qualities that he believes correlate to a high likelihood of success. He has found that determination trumps all other qualities, but frustratingly, it’s the most difficult to discern. “No one, including the people themselves, know how much of that they have,” he says. Graham also looks for the “relentlessly resourceful.” For example, while he was considering whether he should invite AirBnB, the eBay for private travel accommodations, to become a Y Combinator company, the founders told him that they’d designed and sold $36,000 worth of kitschy Obama O’s and Cap’n McCain’s cereal to take advantage of the 2008 election mania. Graham didn’t need to hear another word; he signed them on. Now the company has listings in 8,000 cities in 167 countries.

Every two weeks, Graham and his partners, one of whom is his wife, Jessica, rate their current batch of startups, and they check back regularly as each venture’s story unfolds to assess how good their instincts were. Y Combinator has been criticized for funding small niche concepts with dreams no bigger than becoming a quick acquisition target. TechCrunch editor Michael Arrington riled up the crowd at one Y Combinator event by repeating an inflammatory comment he’d heard from a top VC: “There’s an entire generation of entrepreneurs building dipshit companies to sell to Google for $25 million.” Graham’s response: “Big ideas are not necessarily better than small ones.” He tells his entrepreneurs to think of their starting concept only as a beachhead. “Think of what Microsoft would have been pitching on Demo Day—basic interpreter for a machine when only a couple of thousand [of them] existed.”

Back in 2005, when Y Combinator incubated its first eight startups, Graham couldn’t have imagined handling 15 companies in a year, he says—yet in 2010, Y Combinator has taken on 63. So far, 10 or so of its ventures have produced multimillion-dollar exits, but none have blasted into the stratosphere. As respected super angel Mike Maples says, “the real question is, ‘Can YC become a place that produces the best 10 companies of the year?’ That remains an open question.” Graham isn’t slowing down to wait for an answer. “We’re approaching YC like software,” he says. “You crank up the volume and see where the system breaks.”

Some angels seem to be peeved at Graham’s dominance. They say he’s not as angelic as entrepreneurs believe. To be accepted into Y Combinator, companies—many of whose founders are too green to know better—give up a 2 to 10 percent stake to Graham and his partners. Y Combinator, itself basically the earliest-stage investor of all, with 208 companies in its fold, wouldn’t mind being the next Zynga-size success story, and Graham took on Conway’s SV Angel and ex-Googler Aydin Senkut’s firm as investors. Sequoia Capital also has a stake in the incubator and so presumably gets a first look at companies like Rapportive.

At a secret dinner party in September at Bin 38, in the Marina, virtually all the big-name super angels met to discuss how to keep Y Combinator in check, as well as how to keep deals away from VCs and work collectively (i.e., “collude”) to drive down valuations so each of their funds could own more of a company for less. On hand for Conway’s firm was young partner David Lee, but when this fact ended up in the press and threatened to tar Conway’s reputation, he sent a scathing email to the other participants: “In my opinion, your motives are driven by self-serving factors around ego satisfaction and ‘making a buck.’” McClure responded by parading around in a T-shirt emblazoned with the words “I was at Bin 38 and all I got was a lousy valuation.” What became known as Angelgate showed that the potential for spoils had put an end to the kumbayas, if not to the humor.

Graham thinks that the meeting amounted to little more than a group grumble about the tripling in value assigned to tech startups in the past two years, “the way a bunch of old people get together to talk about their infirmities.” He laughs at the idea that colluding would even work; with all the other potential funding sources out there, the super angels don’t have a monopoly on early-stage investing. “They’d need to be idiots to think that they could actually achieve this, and they’re not idiots,” he says.

It’s unclear where all this energy is taking the Bay Area and the nation. “We are in the fog of battle,” says Stanford’s Blank. Yet it’s hard to imagine that we aren’t riding a bubble of some size when Andreessen’s new firm raised $650 million in three weeks for its second fund devoted to consumer-tech startups, and when John Doerr touts his social-media fund as “a quarter-million-dollar party.” Free-floating money is landing in scarily reminiscent places: an online vintage-clothing boutique and a web company where users post celebrities’ favorite novels and getaway locales.

Graham doesn’t seem to be worried. “After this last YC batch,” he says, “I did think to myself, ‘Could this be a bubble forming?’ because the valuations were so high.” Now he’s convinced that the inflation stems not from anything that angels are doing but from the VCs’ war to get in at the earliest stages. “An investment bubble, in the strict sense, is when people overpay—knowingly—because they think that someone else will overpay later. That’s not what’s going on. This isn’t fools eating up companies that have one quarter of earnings and a vague business model.” Anyway, the overall amount being invested is tiny compared to past booms. When these companies fail, expo­nentially fewer employees, leases, and livelihoods are at stake. 

And that’s the point, Graham says. Small—and relentlessly experimental, quick-footed, and determined—truly is beautiful now. “In the ’20s and ’30s, there was a turn toward big national companies—U.S. Steel and others,” he says. “‘Small’ meant Uncle Joe’s shoe store. Small meant lame.” But it’s clear to Graham that brilliant technologists who work small are anything but lame. Instead, they are the rocket ships everyone is clamoring to board.

Whether the angel-touched companies of the Twitter age will endure, produce something of true value, and propel us all forward remains to be seen. That was the gold standard that the former Silicon Valley—including its VCs—judged itself by. Though Graham won’t say that his entrepreneurs and the angels who watch over them will go down in the history books alongside the giants who created the Valley’s iconic companies, he does speak of working at the center of a shift on the scale of the industrial revolution. Then he checks himself: “Not even the participants know how this is going to end up.”

Diana Kapp is a San Francisco contributing writer.