In 2009, Adam D'Angelo and Charlie Cheever — two former early Facebook employees — were planning to build something new. They had not told anyone about the project, but word about the pair's activity spread — and that got the attention of Keith Rabois, an early PayPal executive.
"I knew both of them back when they were at Facebook," Rabois told BuzzFeed News. "When I heard they were gonna do a company together, I told both of them I wanted to invest. They said 'we aren't telling what we're doing yet,' and I said I don't care."
The site they created, Quora, is now at an important juncture. It has built a large and loyal following within Silicon Valley. And it attracts some of the world's best software engineering and data science talent, outmaneuvering prominent technology companies to get the best young brains into its offices, industry sources have told BuzzFeed News. Once there, they are put to work solving one of the trickiest challenges in computer science: how to train machines to understand the nuance and complexities of human language, and apply that understanding to a giant pool of data to build heavily personalized user experiences.
"We're basically building this map of who the experts are in every given topic of knowledge," D'Angelo said in an interview with BuzzFeed News. "So it's this technical challenge around building out this map of areas of knowledge, and this database of who the experts are, and automatically updating that over time."
But while the technical challenge is great, the company must also solve an equally vexing growth problem. To become truly huge -- like, internet-in-2014 huge -- it needs to break out of its core market of techies and become a mainstream site, as popular among NFL fans as it is among software developers. It's a challenge many Silicon Valley companies have faced in their lifetimes. Facebook eventually had to move away from college campuses. Twitter had to find an audience outside of media and technologists.
The trick for Quora, however, is that it has to ensure that the quality of its content remains high, because the network effects built into a site like Facebook — it's where all your friends are — don't necessarily exist to keep people on the site. To keep people coming back, Quora relies on the strength of its questions and answers and its overall pool of knowledge and experts. D'Angelo's bet is that people's curiosity and desire for knowledge — and desire to share their knowledge — will be just as valuable as the need to connect with friends or stream music.
"Most content startups just accept a reality where Google search or Facebook [News Feed] deliver traffic, and try to play by their rules and optimize their content for search, or for click-bait in your news feed," Josh Hannah, a general partner at Matrix Partners, which has invested in Quora, told BuzzFeed News. "Quora is willing to invest significant time and money on the bet that the best content and the ability to show you just the content you want — both very hard problems to solve — can in the long run build a giant, independent destination."
Historically known as a well of knowledge about Silicon Valley and technology, Quora has since extracted information on more than 500,000 topics from unexpected experts around the world. The list includes former astronauts, NBA statistic nuts, and even an early painter for Walt Disney films. Within the Valley, Quora is considered one of the quietest technology companies, keeping a low profile that is somewhat of a deviation from the industry norm.
D'Angelo and his team of a little more than 100, tucked away on two floors of a small office in Mountain View, Calif., are still in the early days of their site's mission. Quora, in their eyes, has still only amassed a fraction of a fraction of every possible question that needs answering. It has to battle a perception that it's primarily a question-and-answer service focused on the Silicon Valley crowd. And if it's going to become a lasting and vital part of the online landscape, much in the same way D'Angelo's previous company Facebook did, it's going to have to build a business.
With a sizable war chest of financing ($80 million from a round earlier this year), and the ability to potentially tap D'Angelo as an investor if needed (he's personally invested in the previous two rounds and generated a small fortune from his time at Facebook), Quora has risen to become nearly a billion-dollar company, without selling a single ad.
Quora has raised more than $150 million in financing overall, and its investors will say the company isn't as pressured for a sale or initial public offering, and can devote itself to very long-term goals and objectives without having to make money in the near term. Indeed, the board of directors of Quora consists only of D'Angelo and Matt Cohler — a partner at Benchmark, which invested in Quora, who was also an early Facebook and LinkedIn employee.
"The fundamental property is that the same person usually knows about a bunch of different topics," D'Angelo said. "Because people know about lots of different areas of knowledge, it does end up with this natural diffusion process. At this point a very small percentage of Quora is Silicon Valley, and I think it can still spread much further."
As Quora has grown, the challenge is ensuring that the community — often more diverse than what might be found on Wikipedia — adheres to a similar principle of ensuring questions and answers are high quality. "We want the content to last forever," D'Angelo said, with each question finding the perfect answer that will remain evergreen for the life of the site. Other question and answer sites like Yahoo Answers have cropped up, but have become a well of generally useless and low quality answers. (One of the most prominent examples of this is Yahoo Answers' infamous "how is babby formed" thread.)
Quora is often compared to Wikipedia, one of the largest sources of information on the Internet. Despite becoming as large as it has — the English Wikipedia alone generates nearly 10 billion page views a month — Wikipedia is considered a generally reliable source of information thanks to rigorous policing by its community.
"If you look at all these other efforts before Quora, to get knowledge onto the Internet, a lot of them suffered from quality [issues], so that's a big problem," D'Angelo said. "We need to build systems that can automatically figure out what's high quality and what's not, and encourage users to contribute high quality content."
There might seem to be some overlap with Wikipedia, but D'Angelo still considers Quora a source of knowledge for an even longer tail of information on the Internet. Instead of a broad topic — like the International Space Station — Quora's information is by design intended to be more directed and nuanced. Each answer has a real name attached to it — such as when Clayton Anderson, who spent several months on the International Space Station, replied to a question of what it smells like on the ISS. Quora has focused its design on making sure its users are accountable, which in theory drives higher-quality questions and answers.
"A lot of way Quora works is not luck or magic, it's by design," Quora director of design David Cole told BuzzFeed News. "The effect of [requiring a real name] is, all the things that happen you associate with real life, all the things about accountability are then brought into the product itself. I can see that you are a real person, you've had this real experience. I give this example, there's this comment where this guy said, this is an amazing answer, thank you so much for writing this. If you look at YouTube, you would never see something so sincere."
Jimmy Wales, the CEO of Wikipedia, is an avid Quora user. He is also an investor in the company, a signal often seen within Silicon Valley that the services aren't in direct competition. If online commenting on some news stories was representative of the Internet as a whole, readers would "weep for the future of the species," Wales told BuzzFeed News in an interview.
"It's just idiots yelling at each other, so you might think from that kind of a world, 'it's hopeless,'" he said. "And then you go to Wikipedia and say this is amazing, it's this gift from this great group of people. And if you go to Quora, you ask a question and get a lot of thoughtful answers. Any time you can find a way to harness those people, to get the nice people on board with what you are doing, I think it's always interesting."
Getting those people on board and contributing to the site is at the heart of one of Quora's most important technical challenges: ensuring every question finds the user best equipped to answer it. When those two pieces come together, Quora ideally creates a virtuous cycle -- both the questioner and the respondent feel satisfied and engaged -- and creates content that attracts new readers. Those readers will then, ideally, join the cycle, asking and answering more questions.
But getting the right questions in front of the right users is harder than it sounds, and it's at this point that the technical recruiting prowess of Quora kicks in. The company pays competitively with well-funded technology companies with massive profit pools, paying its software engineers an average of $108,470 a year, according to GlassDoor. In comparison, Google pays software engineers an average of $125,295 per year, while Facebook pays an average $122,900, according to GlassDoor. And as always, working at a pre-IPO startup like Quora offers the benefits of getting equity that will boom in value if the company hits the jackpot.
But software engineers and data scientists are already some of the best-paid professions in the country, and salary and benefits can only rise so far until they begin mattering less and less. People also want to work on challenging, meaningful projects, and among developers and engineers the Valley, Quora is seen as taking on some of the most interesting technical challenges of the Internet, combining machine learning with deep levels of personalization and an algorithmic understanding of the quality of questions and answers posted on the site.
"If you work on Quora you usually care about learning, whether it's intellectual or personal development," Quora director of product management Sandra Liu Huang told BuzzFeed News. "People care about the mission but also about learning at multiple levels."
The technical process mirrors that of Netflix and it's recommendation system. Quora has to very quickly discern what a user is interested in, with limited data points to rely on. "No one's gonna sit there and spend hours marking exactly what they want," D'Angelo said.
"It's a lot of statistics and math, and there's a lot of creativity that goes into the ranking of predicting what someone's going to answer," D'Angelo said. That creative challenge is valued by employees. "We don't usually have to worry about poaching," he said.
Beyond the technical problem, Quora also hired an HR manager early in its growth, something uncommon in Silicon Valley companies that are growing quickly but see it as a misallocation of resources. Sarah Smith, the company's VP of recruiting and HR, was brought on as employee number 40, and even as early as 20 employees Quora had instituted a performance review process.
One example of Quora's retention processes is its compensation program, where employees have a ten-year window to exercise their options, instead of having to do so immediately after leaving the company. "People never feel handcuffed, they're here because they want to be here," Smith told BuzzFeed News in an interview. "I'm not a fan of the types of retention tools where people feel handcuffed to their job."
There are other perks that are tuned to keep Quora's employees interested. Across the company, employees are expected to have some programming experience — though, to be sure, not every employee writes code. It also institutes a policy called "continuous deployment," a process where any employee can change the code of the service at any time. At larger companies, the deployment schedule is typically once a week, which could encourage developers to hold off on working on changes to a product's code until the last minute. Such a process also requires extensive care, and the service is loaded with safeguards to ensure any one change doesn't break it — a considerable technical problem on its own.
Collectively, the policies set in place — and the tone, very early set by D'Angelo — are designed to keep employees engaged and satisfied with the products they are working on. When new products or changes are in the works. Quora sets a single person in charge of the execution of that product, giving engineers or designers more responsibility and ownership of anything the company does at a given time (a policy called DRI, or directly-responsible individual) — a structure adopted from Apple, Huang said.
"Sometimes it's an engineer because it's a ranking product — as good as a product manager would be, an engineer is probably just better," she said. "I think that as a construct makes it really clear who is always on point. At Google there's always a little bit of tension between functions, and here we like having that clarity, it helps decisions get made more efficiently and there's less jockeying. We don't like ambiguity where we can avoid it."
Eventually, Quora will have to build a money-making business. And when it comes to the kind of commercial model it may follow, many think Quora's potential as a money-maker looks more like Google than Facebook.
When users are coming to Quora and asking questions, they often carry what is referred to in the industry as "intent." When someone searches for "rain boots" on Google, there is a high likelihood that they are looking for suggestions for rain boots — or, the best case scenario for advertisers, are looking to buy rain boots. Advertisers can then buy real estate that promotes their rain boots for sale.
"Services that have a lot of intent actually end up monetizing really well," D'Angelo said. "There's different ways, exactly, how to make a business about that. There's different ways we can go, we can show ads like Google does, we can have sponsored answers or let people pay for more distribution for their content. But it's pretty aligned with, fundamentally users are using Quora to figure things out and to get answers and to learn."
The challenge — much like Facebook's — will be to ensure that users are not annoyed by advertisements, which would discourage them from visiting the site over and over again. For Facebook, it's a delicate dance to determine an ad placement, and it factors in the quality of the advertisement and how much the advertiser is willing to pay. Inevitably, ads that look more like News Feed stories on Facebook get better placement, as users are less likely to be annoyed by them — and, potentially, much more likely to click on the ad or be encouraged to buy some product.
The precursor for a serious advertising business at Quora will be driving activity on not only on the Web but also for its mobile application. According to AppAnnie, Quora was only ranked among the top 500 social networking applications on the Apple App Store, and has not secured a place as one of the most popular apps. Historically it has not deviated far from that ranking. While app store rankings aren't the definitive metric for success, they tend to be a good barometer for the popularity for an app — for example, Minecraft and Candy Crush saga -- both now worth billions -- still dominate the top of the Apple App Store.
"Marketers are still extremely concerned about scale and reach," Andrew Frank, VP Distinguished Analyst at Gartner, told BuzzFeed News. "Quora may have a very dedicated and loyal following who are high-value customers and are very likely to be influenced by Quora, but as long as their audience is much smaller than Google, or Facebook for that matter, I think they're going to have a tough time competing in quality alone. I think scale is still really important to advertisers."
But D'Angelo's goal is still to grow its user base methodically, rather than virally, to ensure the quality of answers remains high. Quora requires a user to sign up with a Facebook account or email address to actively browse the site. That's because Quora needs at least a few data points if it's going to offer a good set of content to new users and keep them interested, instead of confused, as to what the purpose of the site is.
Beyond the technical problems of personalization is a broader one: international expansion. Already, Quora is attracting activity from outside the U.S. — though the U.S. is still where the majority of its users are — and to cater to those new, potential mainstream users it will have to customize the service. Part of that is ensuring that it supports multiple languages, and partis integrating well with popular social networks abroad. For example, Quora would have to integrate with vKontake in Russia, or chat networks like WeChat in Asia.
Quora started as a niche product specialized in Silicon Valley and technology — but so, too, did many other services that have gone on to be massively successful.
"All the services that have become very large and successful started from a position of strength in some smaller, defined area and grew out from there," said Matt Cohler, a general partner at Benchmark — which led the early investment in Quora as well as in Uber, Snapchat, Twitter and Instagram. "That was true of LinkedIn too, which has obviously had a long phenomenal journey over the eleven and a half years that have transpired since then. We've seen the exact same thing happen with Quora, it's always happened that way and it continues."
While the company doesn't disclose metrics, it does say its usage and user base has continued to grow three-fold every year, as it has for some time.
"Honestly, we have our hands full," D'Angelo said. "It keeps growing, that's challenging for us to keep up with that. It's probably going to continue next year. A lot of our energy goes toward making the system stay really high quality as we get to being a product that the whole world is using."