For a guy running such a beautiful website, Ben Silbermann looks like hell; he has prominent bags under tired, watery eyes; his shoulders hang heavy; his shirt is wrinkled; and his dark hair is uncombed. When he speaks - with the open-vowel inflections of his upbringing in the United States Midwest - his voice is so slight that it often gets lost beneath the din of other conversations. When he moves, it is with the economy of a marathon runner trying to conserve every last bit of energy on the eve of a big race.
"I'm tired," says the 30-year-old chief executive of Pinterest, the social scrapbook that's one of the hottest websites on the planet, as he prepares to shovel down a bowl of noodles a few feet away from his desk. Silbermann leaves for the office at 7am most mornings and works non-stop until dinner. His only respite, if you can call it that, comes in the predawn hours, when he takes his newborn son, Max, into his arms and fires up his laptop to check his e-mails. Just a few weeks before Max was born, in early July, Silbermann declared a companywide lockdown, ordering his 35 employees to come early and stay late in order to build new iPad and Android applications. The goal: to stoke growth. He ordered commemorative T-shirts with the phrase "Summer of Apps" printed across the chest, and he cut off almost all contact with anyone outside the company, including potential business partners.
Such is life at what earlier this year was declared the fastest-growing web service in history.
This schedule is the price of running a site as beloved as Pinterest, which has won fans thanks to its breakthrough design. Think of Pinterest as a giant digital catalogue filled with the web's most beauti-ful images. Users - "pinners", in the twee parlance of the site - copy images they find elsewhere on the web and store them on one of their personal Pinterest pages, called pinboards. (Everything in the company's office, down to the Wi-fi password, has the word "pin" in it.)
On first blush, Pinterest may sound like a hundred other social media websites where people share images and comment on them. But the design choices of Silbermann and site co-founder Evan Sharp, based on a new way of browsing that dispenses with the web's rigid rules of presenting content, have made the service incredibly addictive. To create a pinboard is to say to the world, "Here are the beautiful things that make me who I am - or who I want to be." Young women use Pinterest to plan their weddings, men collect watches and bikes in de facto gift registries, and couples assemble furniture sets for their new homes. Pictures of attractive men and women in various states of undress abound. The sum of each user's choices is displayed in an ever-changing pastiche on each person's homepage.
"When you open up Pinterest," Silbermann says, distilling his vision, "you should feel like you've walked into a building full of stuff that only you are interested in. Everything should feel handpicked for you." In other words, it's a store in which every single product has been tailored to your needs, ambitions and desires.
Pinterest's combination of elegant design and smart social networking dynamics has made it the internet's latest Great Revenue-Generating Hope, at a time of extreme scepticism about newfangled ad-based business models. While Facebook's one billion users seem stubbornly resistant to the entreaties of marketers, the Pinterest set has already been opening their wallets.
"This is going to be an alternative form of e-commerce," says Greg Fant, chief marketing officer at One Kings Lane, a design-centric home-decor web retailer. Pinterest users who find their way to One Kings Lane are three times more likely than the average One Kings Lane shopper to make a purchase; referrals to One Kings Lane from Pinterest have nearly surpassed traffic from Facebook. Across the web, the average sale resulting from a Pinterest user following an image back to its source and then buying the item is US$180, according to research from e-commerce firm RichRelevance, compared with US$80 for Facebook users and US$70 for Twitter users.
Pinterest gets none of that. The company is in what's known as a pre-revenue phase. But that doesn't mean Silbermann hasn't thought about how Pinterest is eventually going to get its share of all that e-commerce.
"We talked about it a lot," says an early employee who asked to remain anonymous because he was flouting Silbermann's wish to keep the focus away from dollars and cents. "There was never a doubt in our minds that we could make a s***load of money."
SILBERMANN AND I MEET midway through the Summer of Apps. The company has just moved into a loft in San Francisco's SoMa district, and Silbermann is negotiating to lease an additional 60,000 square feet next door to accommodate hundreds of new employees.
"It's a busy time, but it's good," says Silbermann, a sensitive man with a boyish face and a gentle demeanour who couches even his harshest criticisms in a sea of qualifiers and who often uses the word "gosh" without irony. As in, "I'm not going to lie. There are really stressful days when I feel like, gosh, I just want to make the most of this."
What Pinterest has done is solve the problem of discovery on the web. And it has been a problem for a while. Say you want to buy a gift for your mother. A search for "nice Mother's Day gift" on Google - or even "very special, very expensive Mother's Day gift" - isn't going to be much help. Google depends on finely tuned queries in order to yield useful results. (This makes Google a great advertising platform, because it delivers customers who have already expressed an intention to buy something.) But talk to Google the way you might talk to a clerk in a department store, and it won't know where to begin. You could go to the web's superstore, Amazon, but that's not much better.
"You spend three hours buying a US$20 toaster," says Barry Schwartz, a psychologist and the author of The Paradox of Choice.
"Browsing in e-commerce is a more difficult problem than search," says Leland Rechis, a director of product experience at Etsy, which sells handmade goods. "Amazon and Google pretty much stink at browsing."
These failings of the two great internet moneymakers are not entirely their fault. The structure of the internet is partly to blame. The web was laid out not in human terms but as a series of ever-more-specific menus, making the kinds of free-associative leaps that routinely happen in a shopping mall nearly impossible. Contrast this with the fact that many of the world's most profitable retailers - Apple, for instance - succeed by curating a smaller selection of products for their customers.
Pinterest is the first site to bring that curated vibe to its user experience. Schwartz decries how the web exacerbates the choice problem, but he gets excited when Pinterest is mentioned.
"People who can afford it shop in boutiques. Pinterest strikes me as the same, except that the selectors of the merchandise are your friends. I'm not surprised it's exploding."
BORN INTO A FAMILY of doctors in Des Moines, Iowa - his parents run an ophthalmology practice, and both of his sisters are physicians - Silbermann went to a public high school where he was a nationally competitive debater. As a high-school junior, Silbermann earned a spot at the prestigious Research Science Institute, a free academic programme affiliated with the Massachusetts Institute of Technology and the US Department of Defense that is perhaps the country's most selective summer camp. He figured he'd be a doctor as well.
Silbermann was accepted into Yale university and spent the next two years prepping for the medical-school entrance exams. Then he changed his mind. He bought a leather portfolio, put together a résumé and spent his last year of college applying for consulting jobs. After graduation, he found himself at a Washington DC management consultancy.
He tried to be a good consultant, but his interest drifted again. It was the mid-2000s, the beginning of the social web boom, and Silbermann began obsessively reading tech blogs and following the start-ups of the day - Digg, MySpace, YouTube and Facebook.
"So many things that I was excited about as a kid were about proximity," he says. "The idea that somebody could grow up in rural Iowa and be into break dancing because of YouTube - that was a really simple, profound idea."
The more he read, the more Silbermann became convinced that he was missing the story of his generation. His peers were building world-changing companies. What was he doing?
"It was this burst of frustration," says Silbermann.
He and his girlfriend quit their jobs and moved West. At the end of 2006, he landed a customer support job for Google's advertising division - a minor miracle as far as Silbermann was concerned, considering he had no technical background. He spent two years at Google before striking out on his own in 2008.
Silbermann's first product, which he launched with a college friend, Paul Sciarra, was a shopping app called Tote. The partners managed to raise a little seed funding, and Silbermann spent a year refining the app. Tote failed to take off, but it did reveal something interesting: rather than buy things, people used Silbermann's app to e-mail themselves pictures of products to view later.
This struck a chord with Silbermann, a life-long collector.
"I really liked insects," he says. "All kinds: flies, grasshoppers, weevils." He'd pin them to cardboard and put them in little shadow boxes.
In the autumn of 2009, while Silbermann was still struggling with Tote, he had a drink with Sharp, a student at Columbia University's Graduate School of Architecture. They bonded over their mutual love of infographics and then they started talking about collecting. Silbermann suggested that a digital collection - of books, clothes or even insects - could be a powerful medium for self-expression. Silbermann abandoned Tote, and he and Sciarra began working with Sharp on the new idea.
"Most people don't have anything witty to say on Twitter or anything gripping to put on Facebook, but a lot of them are really interesting people," Silbermann says. "They have awesome taste in books or furniture or design, but there was no way to share that."
As they sketched out Pinterest, Silbermann and Sharp cast aside many then-predominant orthodoxies of web design. Facebook, Twitter, Tumblr and other content-driven websites were based on "feeds", lines of text or images that ran from top to bottom by time. They wanted to create a design that allowed users to browse multiple images at once.
"We were really excited about bringing something that wasn't immediate and real time, something that wasn't a chronological feed," says Sharp, who is serious, bearded and armed with a history degree from the University of Chicago. The idea was to remove the rigid organisational strictures that the web imposed - directories, time stamps, pagination - and replace them with a grid of images that would feel more like visiting a store or a museum.
The trio spent four months batting versions back and forth. "This was the first one," Sharp says, pulling up a grid of images, the borders of which turn a garish shade of royal blue when a mouse hovers over an image. "I must have been trying to give Ben a heart attack.
"From the beginning, we were aware that if we were going to get somebody to spend all this time putting together a collection, at the very least, the collection had to be beautiful," Silbermann says.
As Sharp puts it, "The grid was everything." His final version displayed interlocking images of a fixed width and varying heights. When new images were pinned, the entire site would rearrange itself, meaning users rarely saw the same homepage twice.
Pinterest's design overturned conventional wisdom in a number of other surprising ways. Silbermann rejected the idea that entrepreneurs should hack something together, open it up to the public and tweak, iterate and pivot. At a time when "gamification" was hot, Silbermann and Sharp declined to feature any elements that might encourage pinners to compete with one another. Pinterest offered no leader board or any other way to find the most popular pinners and, unlike Twitter, which makes a person's follower count the focal point of the site, Silbermann did not include it on the homepage.
Most web companies design their sites to generate page views, which helps them show momentum to investors and, in theory, makes them more attractive to advertisers. This often means producing extra page views by any means necessary. A central element of Pinterest's design, though, was the then-novel "infinite scroll", automatically loading more images as the user expands the web-browser window horizontally or goes towards the bottom of the page. The decision meant that Pinterest would generate fewer page views than most websites, but it also meant users spent almost no time clicking buttons or loading pages.
Pinterest's traffic has come from demographic groups not normally associated with fast-growing websites. Nearly 80 per cent of Pinterest's users are women, most between the ages of 25 and 54, according to Google DoubleClick Ad Planner.
Silbermann and Sharp's grid has since proliferated across the web. It has inspired numerous copycat sites that mimic its look and feel down to the font selections.
After Pinterest went live, in March 2010, Silbermann spent the next 12 months designing the character of its community. The grid itself is only as beautiful as the pictures that fill it, and Silbermann worried that Pinterest would be inundated - and rendered useless - by party pictures, porn and all manner of internet weirdness.
To prevent the site from turning into another online photo album Silbermann prevented early users from uploading pictures directly from Facebook, and he initially limited sign-ups. Many of the early invites went to a group of design bloggers whom Silbermann recruited personally. He gave each user a limited number of invitations, urging them only to invite others whose taste they respected.
In Pinterest's early days, Silbermann gave out his mobile phone number, attended blogger meet-ups and composed weekly e-mails that were sent out to Pinterest's tiny, but growing, community.
"It's like you've built this little city with nobody inside of it yet," he says. "You want to fill it up with the right kinds of people who are going to teach future people what they should be doing when they move in."
Most Silicon Valley types look at early users as viral marketers; Silbermann saw them as role models. (Until recently, Pinterest's welcome e-mail advised users to "pin carefully" because "your pins set the tone for the community".)
Silbermann's view of design as social engineering is on display when I sit in on Pinterest's weekly design review.
"We start with what we want people to feel and what we want them to do, and there's a design that feeds into that," he says.
The focus of the two-hour critique is the company's forthcoming iPad app, which Silbermann believes is critical to Pinterest's ability to change the way people discover things. "You can use a tablet in places and at times when your mind is free to explore stuff - like having a beer or sitting on the couch," Silbermann says. "Pinterest was made for tablets."
The app includes a feature called "sheets", which allows users to trip through Pinterest pages in the same way one might bounce around a shopping mall. (The iPad app was No1 in the App Store within a day of being released.) With a finger flick, users can quickly cycle back and forth through pinboards, user profiles and pages for topics such as gardening or travel without ever returning to the home screen.
"Most apps assume you have this navigation tree," says Sharp. "But that's not the way you discover. It's not the way you shop. It's not the way you go through a museum. That's our interaction model."
For all the talk of Pinterest's promise as a business, it is generating no revenue whatsoever. The fact is far from a secret. Go to its website, click on "Help", and then click on the frequently asked question, "How does Pinterest make money?" The answer is, in short, that it does not.
Investors seem happy to let Silbermann take his time to develop a business model. To date, Pinterest has raised US$138 million from an A-list roster of investors. In May, Pinterest raised US$100 million from a group led by the Japanese e-commerce giant Rakuten, a funding round that valued the company at US$1.5 billion. Rakuten's founder and chief executive, Hiroshi Mikitani, told The New York Times that if Pinterest had been for sale, he would have bought the whole thing.
Some of Pinterest's most popular users - and the brands that want to reach their followers - are not waiting for Silbermann to create commercial opportunities. For example, certain brands pay Satsuki Shibuya, a 31-year-old designer with more than a million followers, between US$150 and US$1,200 every time she pins an image of their product. She refuses to say which brands she's worked with and, because paid pins look identical to unpaid ones, it's hard to guess.
"A lot of brands are getting into the game," she says. "It's a smart move. They're already putting ads in magazines and there are 10 times as many people looking at Pinterest." Earlier this year, she said no to an advertiser because the items seemed too far removed from her personal style. "For three or four pins," she notes, "I could have bought a car."
"The power pinners are really wrestling with what to make of this opportunity," says Su Wu, who works in the research communications department at the University of Southern California and, as a sideline, pins to an audience of 300,000 followers. One notable exception: Jane Wang, who has the largest follower count on the service - 3.4 million - and happens to be Silbermann's mother.
"My mom does not get paid to pin things," Silbermann says.
She probably gets offers, though. Fant of One Kings Lane says that earlier this year, his company began experimenting with paid-for pins, offering 100 of the top home-furnishing pinners a substantial percentage of the proceeds their pins generate.
"We're still figuring out how best to scale the programme," he says.
So what happens when amateurs who are not Silbermann's mother are offered fat cheques from marketers? Should Pinterest try to stop it?
Silbermann stares at me blankly. "We're just trying to preserve the authenticity," he says. "Pinterest is a young platform."
It's sometimes hard to tell whether Silbermann is being sincere or coy when he says things like this.
This year, Silbermann hired Tim Kendall, an engineer with an MBA from Stanford University, to serve as the company's head of product management and partnerships. Kendall is credited with creating Facebook's monetisation strategy in 2006, and he helped Tumblr develop its recently launched advertising platform.
"What's unique about Pinterest is that as you go through these streams of content, some of that content will be purely organic and non-commercial and some of it will be commercial," says Kendall, "and you're not able to discern between the two.
"That's pretty awesome for future business potential."
However, as Kendall suggests, selling products might be just the beginning for Pinterest. Jess Lee, chief executive of fashion-focused Pinterest rival Polyvore, says, "E-commerce 1.0 was driven by digital cameras, electronics, hard goods - things with technical specs that you can plot on a chart. But why does someone pay US$25,000 for an Hermès handbag? It's not rational. It's based on taste; it's totally amorphous and not trackable. That's e-commerce 2.0."
Polyvore, which is profitable and attracts five million users a month in the US, makes about half of its revenue from taking a cut of what its users discover on its site and end up buying elsewhere. The other half of Polyvore's revenue, though, comes from magazine-style advertising campaigns by the likes of Nordstrom and Mercedes.
"I want the entire sales funnel," says Lee, meaning both the ad revenue that creates the desire and the sales that fulfil it.
A clue that Pinterest might be aiming for this approach: the investment by Rakuten, a US$5 billion online retailer.
"We share a vision with Pinterest," says Mikitani. "Our theme is that shopping is entertainment. If everybody bought everything on Amazon, it would be a dystopia."
Pinterest could continue to refine the way it encourages the discovery of products - incorporating splashy (and expensive) ads - and it could sell goods on pinboards on behalf of retailers. Or it could turn itself into a giant affiliate marketer. Whereas electronics, the bread and butter of most e-commerce sites, has painfully narrow margins, the categories that dominate Pinterest, such as apparel and home furnishings, can have initial margins of 50 per cent or more.
I try some of these theories out on Silbermann, who demurs, obfuscates, cracks wise and generally prattles on about the need to focus on his product. Perhaps out of frustration, I suggest that perhaps he hasn't yet figured out how his beloved grid will actually make money.
Silbermann smiles, looks me in the eye and ever so politely says: "We have a pretty good sense.