The Fall of Tumblr: How a Generation of Innovators Remade Online Social — and Then Faded | QDaily Business History

Facebook and Tencent had the last laugh, but far more than those two companies shaped how you socialize online today.

The re-election campaign was about to begin, and the American president, now in his third year, was as bold as everyone expected.

To win over young voters, his Twitter account wasn't enough. He fired up Tumblr and rolled out a slogan: "Tumbling for Change."

That president was Barack Obama. In 2011, the third year of his first term, he opened a Tumblr account to court young voters for his re-election bid. That June, 14 million people opened Tumblr in the United States alone. For comparison, Instagram had just 5 million users worldwide at the time.

Tumblr didn't yet carry its Chinese nickname back then; it was the biggest social-media startup since Facebook and Twitter. Its founder, David Karp, was only 20, the next young man after Zuckerberg and Jack Dorsey to command Silicon Valley's full attention.

Tumblr took its name from "Tumblelog," whose Chinese rendering is "light blog." In 2005, a 15-year-old German prodigy, Chris Neukirchen, built the world's first light-blog product and named it by joining "Tumble" (to scroll) and "Web-logging." A year later, Karp tweaked the word and launched his own light-blog product: Tumblr.

The light blog was once seen as the future of online social networking. Chinese internet entrepreneurs fired up their photocopiers. In 2011, a wave of imitators appeared one after another: Diandian, Sina Qing, NetEase LOFTER, Renren Xiaozhan, and more.

Today, whatever you make of Donald Trump, it's hard to imagine him choosing to govern by posting on a Tumblr that has just finished a porn purge.

The mainstream had moved to Facebook and Instagram. By January 2017, a Bell Labs paper posted to arXiv found that one in four Tumblr users was there for adult content. An app that had once drawn even Obama as a campaign tool had drifted to the margins, surviving on smut.

For a Tumblr that had already gone bad, the purge was suicide. The company had long stopped releasing monthly active user numbers; by eMarketer's count it now stood at about 21.3 million, barely moving in years. Meanwhile Instagram — which trailed Tumblr back in 2011 — now had more than one billion monthly active users.

In truth, the purge only sped up a decline that had begun much earlier. By post volume, daily updates on Tumblr started climbing in 2011 and peaked at 106 million in 2014. After that, user posting fell straight down — three years later it had dropped back to 2011 levels.

In 2012, a confident Karp told Britain's The Guardian: "We're now thinking about which company to buy, because I hear AOL is cheap." Today the Yahoo that bought Tumblr has been folded together with AOL into a Verizon subsidiary, and Tumblr is just a trivial piece of Verizon's vast business. News of the purge briefly dragged Verizon's share price down, but it bounced back within three days. Next to Verizon's telecom-carrier business, whether Tumblr lives or dies hardly matters.

Another company back in the public eye on the strength of bad news is Pinterest. In 2018 its user base grew 27 percent year over year, reaching 250 million monthly active users. But its ability to make money still hasn't been proven.

Since 2017, Pinterest has kept talking about going public, yet its revenue has never hit expectations. For a company that has raised more than $1 billion, carries a valuation above $11 billion, and has yet to prove it can turn a profit, an IPO is the only way for investors to cash out and for the company to keep running. Its imitator Mogujie has already gone public on a photo-social-plus-e-commerce model, yet even with the giant Tencent behind it, revenue hasn't budged.

"Pinterest really is a unicorn — it just doesn't look like one," The New York Times observed.

Snapchat, which listed earlier and was once seen as Zuckerberg's challenger, is now working hard to prove it isn't a bubble. In the two years since its IPO, the company's share price lost three-quarters of its value, falling from the offering price of $27.09 to $6.18. If Wall Street can't read young people, three straight quarters of falling active users say Snapchat can no longer hold onto them either.

Weibo, which listed earlier still, watched its share price slide for a full year, then capped the year by announcing it would cap the displayed retweet count at one million and stop revealing more precise figures. Twitter's decline has been just as severe and just as relentless.

New attempts keep coming. On January 15, Toutiao, Smartisan, and Wang Xin — fresh out of prison — each launched their own social product on the same day, though probably no one believes any of the three apps can really challenge WeChat.

Over the past six years, social networking on the mobile internet has largely settled. Facebook and Tencent, these two companies, hold a firm monopoly over our social relationships.

But when it comes to how we socialize online, the credit belongs to far more than just Zhang Xiaolong and Zuckerberg.

How did social networks become what they are today? Two companies didn't decide that.

Ten years ago, global smartphone shipments were still surging. By the end of 2008, four billion people around the world had a mobile phone.

The modern internet social world had already begun to surface. The beeps in small-town internet cafés never stopped, and boarding-school students who didn't yet have phones slipped out over the school gate after midnight to chat with QQ friends they'd never met. That March, QQ's peak simultaneous users passed 40 million.

On the other side, Facebook — born in a Harvard dorm — drew huge numbers of users to upload their own information and build their own profile pages. They submitted their photos, names, ages, occupations, education, likes and dislikes, contact details, and locations. Facebook strung those individual pages into a network, and the first social network was born.

But for most of those four billion people, the social network on their screens was still the phonebook and the text message. The link between two phone numbers worked just like the user IDs of later social networks, and people sent messages of no more than 160 English characters or 70 Chinese characters.

Texting wasn't, as it is now, just a slot for verification codes and spam ads; it was a serious social tool. In 2005, China sent more than 200 billion text messages in total — 217 times as many as five years earlier. Major TV programs regularly brought in experts to discuss "the harms of texting too much," much as we talk today about the harms of social networks.

Ten years later, social was the business of two companies:

Apart from Douyin, of the social apps with more than 500 million users, four belong to Facebook (Facebook, Instagram, Messenger, WhatsApp) and two belong to Tencent (WeChat, QQ). Then again, it's hard to say how much Douyin has to do with social at all; it's more a video-entertainment platform, like YouTube.

Tencent and Facebook may dominate everything now, but the move from texting to WeChat and Facebook in fact gathered up the bursts of inspiration of at least dozens of entrepreneurs and the work of thousands of engineers and product managers exploring one social feature after another. The names buried in the rush of change together decided how we socialize today.

1. How to build a social network fast: WhatsApp and Kik wrote the template for everyone

In 2008, the colorful autograph books that Chinese middle-school graduates passed around had a dedicated line for friends to write down their QQ number. Punching in that string of digits was how you found and added friends on QQ. That was how you built a contact list on an early social app.

One string of digits, one ID — that was the basic setup of an early social network. But it changed fast.

In January 2009, Jan Koum, just out of Yahoo, bought his first iPhone and realized that the App Store — open for only seven months — was going to upend the entire software industry. On his first startup, Koum spent a few days paging through Wikipedia's directory of international dialing prefixes and used it as the foundation for WhatsApp's first lines of code. Then he spent months entering hundreds of area codes so that users could log into WhatsApp with their phone number and send free messages to any other phone number worldwide.

The real opening came later that year, when Apple introduced push notifications, letting apps receive messages at any time — before that, an app had to stay open in the foreground, which made it all but useless for communication.

WhatsApp 2.0, which supported push notifications, launched at year's end. The people it drew in found that once logged in they could import their phone contacts directly, see which friends had also downloaded WhatsApp, add them with one tap, and trade free messages right away.

For the first time, the familiar faces from your phone's address book were carried into a social app. WhatsApp dropped the user-ID design and turned each person's phone number — which everyone already had — into their WhatsApp account, doing away with the hassle of typing in an ID to search for friends.

A year later, Ted Livingston, just out of an internship at BlackBerry, noticed WhatsApp, which by then had 250,000 users.

"At the end of 2009, I saw WhatsApp making a huge splash, but their service was iPhone-only at the time. So I went to BlackBerry's people and tried to talk them into turning BBM into a cross-platform service to compete with WhatsApp," Livingston recalled — and he was turned down. So he built Kik Messenger himself, and the world's first cross-platform free messaging app went live.

On adding friends, Kik went a step further than its predecessors: besides letting you add contacts from your address book, it cross-referenced different people's contact lists to suggest "people you may know."

The feature was quickly copied by social apps the world over and became basic equipment for any app with even a trace of social function. In a 2013 interview, Livingston called it "a feature every app has already copied."

Tencent, a Kik investor, was no exception. At this year's WeChat Open Class, Zhang Xiaolong brought up Kik and told the story of how it sparked his idea for WeChat.

2. The News Feed: how Facebook got billions of people to happily waste all their time staring at a screen

The first Facebook was born on the PC, so early Facebook carried the style of the PC-era internet: open one person's page, close it, then open another. That was what a social network looked like in those days.

If early Facebook was like everyone keeping an unlocked diary, the arrival of News Feed was like gathering everyone's diaries together and printing them in a newspaper with no last page.

The new feature unsettled users — people were more sensitive about privacy back then. But Facebook held the line, and its user count doubled in the process. By constantly changing its algorithm and adding interest-based recommendations, Facebook got more and more people in the habit of using the feature.

As it turned out, the News Feed not only raised user engagement and time on the platform — it also created far more ad slots. On News Feed's tenth anniversary, Zuckerberg said: "News Feed was one of the biggest bets we've made over the past 10 years. It helped shape our community and almost the entire internet."

To this day, Facebook is still the most commercially successful of all social apps. In the third quarter of 2018, Facebook's advertising revenue was $13.539 billion, up 33 percent from $10.142 billion a year earlier and accounting for 98.3 percent of total revenue. Of that, mobile advertising made up about 92 percent of the quarter's ad revenue.

Today the News Feed is basic equipment in nearly every app. Toutiao, Dianping, Douyin, Zhihu — almost every content app you use regularly is tied to it.

3. Tumblr put the blog on the phone, and for a while it looked like everyone would join in creating

In March 2005, Chris Neukirchen, a 17-year-old German student, built the world's first light-blog product, Anarchaia, and named this new kind of blog "Tumblelog," a word made from "Tumble" (to scroll) and "Web-logging."

The name itself said what a light blog was: scroll down, see more posts.

In February 2007, the 20-year-old David Karp, working in his mother's apartment, built his own light-blog product, tweaked the word "Tumblelog," and named it Tumblr.

On Tumblr, users could post videos, audio, text, social bookmarks, photos, and even other people's blog posts straight to their own blog — no need to upload to YouTube, Delicious, or Flickr first, and no need to set up your own WordPress database.

"Why is traditional blogging boring?" Karp said. "Because not everyone likes to express their ideas in words."

Beyond adding more media formats to the blog, Tumblr made writing simpler. "When you write on a light blog you don't need context," Karp said. "The blog itself is your context." Scrolling down to keep pulling up more of a blogger's older posts felt like going to Weibo to see what someone had posted before — except Tumblr got there much earlier.

"When you read a person's past writing in a feed, it's like opening a fascinating window into their life. You get a very vivid impression of a real person," Karp said.

In Karp's original vision, the "lightweight blog," freed from the burden of long-form writing, would encourage more users to create. And at first it really did go that way. In April 2009, the platform took in 40 million active visitors, with a user retention rate of 85 percent. Of the one million blogs at the time, most were actively creating, with 80,000 to 100,000 new light-blog posts appearing each month.

But in the end, once the mobile internet reached everyone, consumers far outnumbered creators — and that is what the world really looks like.

4. Instagram's square photos plus filters: getting everyone willing to share pictures

Taking a presentable photo was never easy, even though camera phones had been common since the 1990s. But the real ignition of the photo craze starts from 2010.

The sharing problem had long been solved — QQ and Facebook had supported photo sharing for years. But it was still genuinely hard for an ordinary person to take a presentable photo. Whether the camera's usual 3:2, the phone's 4:3, or 16:9, the frame is a rectangle. A rectangle calls for composition.

Even learning the rule of thirds would make your photos much better, but for a broad market of four billion people that was still not simple enough. So Instagram lifted the square viewfinder of the Polaroid camera and, from its 2010 launch until 2015, restricted every shared image to a square.

A square photo only asks the photographer to put the subject in the center — no more composition pressure. And square images also made Instagram's layout look tidier.

Filters served much the same purpose: letting users who weren't much good at photography touch up a shot at the lowest possible cost. Bad lighting, even a shot out of focus — none of it mattered. Just aim, snap, pick a filter, and in one second rescue the photo into something that looks decent enough.

Instagram's design encouraged more users to bravely post their photos. In its first five years, users uploaded 40 billion photos to Instagram. Flickr, which leans more toward photography enthusiasts, took until its eighth year to reach 8 billion.

The square came to be seen as the shape best suited to social spread — even the iPhone's built-in camera added a square shooting mode. And today, when you post a set of photos on WeChat, they too get shrunk into little square thumbnails.

At this year's WeChat Open Class, Zhang Xiaolong explained the Moments design that encourages photo posting: "I think for most people, writing a passage of text is a hard thing to do. Especially when that passage also has to present your persona well. But posting a photo is much easier."

5. Retweet, Like, Pin: creating content gets reduced to taking a position

Tumblr proved how hard it is to get people to create content on the mobile internet. Twitter, born in the same era, had come to the same realization.

When Twitter launched in 2006, it was described as "broadcastable text messages." The first Twitter was designed around the SMS system, so, like a text message, it capped each post at 160 characters. Twenty characters were reserved for the username, leaving 140 to write with.

This seemingly halfway approach actually lowered the barrier to publishing and stimulated more people to create. A mere 140 characters meant that even someone with nothing much to say could post something for the world to look at — and it was clearly better suited to the mobile internet than blogging.

But that still wasn't enough to sustain most people's urge to create. The feature that really stimulated sharing was the retweet — the "RT" — adding the two letters RT in front of someone else's post to join in passing it along. Soon you didn't even need the two letters: Twitter added a one-tap retweet in 2009. Rather than post more original content, jotting a few characters of opinion on someone else's words — or just retweeting outright — became the norm.

Twitter was catering to an ever-lazier phone user, and other social apps were doing the same. The very year Twitter shipped the retweet, Facebook shipped the Like (that famous thumbs-up).

From February 2009 on, content a user liked on Facebook would be recommended to their friends — in effect letting one group of people pick content for another.

Pinterest went further still, building its whole service around the "Pin": letting users conveniently file away images they had made or collected from others into a system of their own (Boards), quickly taking possession of others' content and lowering the barrier to gathering information almost to zero. At the same time, what you pinned was seen by other users — in effect a free round of distribution for that content.

For the platform, the Pin let a limited pool of content be reused as fully as possible, easing the pressure from users' fading appetite to produce while feeding their appetite to quickly consume what others made.

6. Snapchat killed the shame of posting video — and became the most-copied company of recent years

If getting users to post photos already took product managers scheming to nudge them along in secret, getting them to post video was harder still, because the bar for shooting a good video is a hundred times higher than for taking a good photo.

Snapchat was a project Evan Spiegel and Bobby Murphy made for a class assignment in a Stanford product-design course, and it drew no media attention on launch day. But just a year later, Zuckerberg went in person to talk acquisition with Spiegel.

Snapchat is the true owner of a string of important social innovations. Its design has long been knocked as immature and clunky, and its disappearing-message mechanic blamed for fueling child pornography. But several of its designs have been copied wholesale by today's biggest social networks.

Its most central invention was the "disappearing message" mechanic. Before Snapchat, every social app was straining to help users save their data.

In 2012, Evan Spiegel wrote up Snapchat's design thinking: most photos on photo-social apps right now are polished and stylish, but we want to create a place for young people to share awkward selfies and funny pictures with friends.

That design brought the bar for video sharing crashing down at a stroke. Square photos, filters — none of it beats the relief of content that vanishes on its own. The psychological pressure of sharing had been cut to the minimum; Snapchat's idea wasn't to help creators prettify their work but to make them stop caring whether it looked good.

Snapchat's other big innovation was Stories. In essence it was a video feed that auto-destructed after 24 hours; you could post photos, videos, and GIFs and add animated effects and stickers. It meant a user could share photos and videos not only with individuals but into their own Stories, for friends to browse in the open.

"What we want to do is let a person truly record what they're going through, and then let their friends see it. What this promotes is authenticity, not beauty."

That is what Zhang Xiaolong said on January 9, 2019, explaining WeChat's newly launched video-status feature. He spent several minutes laying out every detail of WeChat's video status — details that line up essentially one-to-one with Snapchat Stories, from the timed disappearing to the ring around the avatar. China's greatest product manager explained Snapchat's features start to finish without ever mentioning its name.

And it isn't just WeChat. Stories has been copied across the board by Facebook, Instagram, and WhatsApp, who didn't even change the name. By April 2017, Instagram announced that 200 million people were now using Stories every day.

7. Moments: WeChat starts handling every social need in a single app

WhatsApp was an app for staying in touch with friends; Facebook and Instagram were apps for putting opinions and status out to the world. WeChat fused all three, letting users stay in touch with friends while also putting their opinions and status on display to the outside.

WeChat caught fire early as a Kik imitator, but had it stayed a chat feature it would not be the WeChat of today. When you're not talking one-on-one with a friend, part of you wants everyone to take in your persona. In other words, beyond an IM feature, WeChat needed a public square.

In April 2012, WeChat launched Moments. Of its billion users today, 750 million open Moments every day.

Moments built users a social network for showing themselves to everyone they know, satisfying the urge to "set up a persona." WeChat's other huge change, official accounts, also reaches most users by way of Moments.

And Moments has kept changing — from its early defaults of posting photos, conversations visible only to mutual friends, and images auto-thumbnailed until tapped, to features like "visible for three days" and "visible for six months," rolled out so users wouldn't feel the strain of more friends and more eyes.

You could say that the ecosystem Facebook built across three apps with more than $10 billion in acquisitions, Tencent pulled off with WeChat alone — because of Moments.

Building a product and running a business are two different things

In January 2011, a Chinese team released TalkBox and piled up a million users in under three days. The app had basically one function: hold down the screen to record and send a voice message. Four months later, WeChat shipped an identical feature, and two years after that, TalkBox was dead.

Guo Bingxin, TalkBox's founder, would later talk about it with indignation and regret. But TalkBox in the end had built only a feature, while its rivals had built complete products.

And even a good product wasn't enough.

The social apps above, each with more than 100 million users, weren't just good products; their founders were almost all first-time entrepreneurs with a genius for product. Kevin Systrom and Mike Krieger, who built Instagram on their first try; David Karp, who built Tumblr alone in a small apartment at 20; Zuckerberg, who founded Facebook in a college dorm; and, a decade later, his younger rival — Snapchat founder Evan Spiegel.

The products they built for their peers caught on, and what mattered more than experience was instinct and empathy.

But in the end the evidence showed that simply building a good product wasn't enough. You also had to want to run a good company, ready from the start to make money. WeChat, of course, was an exception: with Tencent behind it, WeChat could go years without worrying about money and just think about making the product better.

"Not knowing how to use a product to make money" also explains why Tumblr, Pinterest, Snapchat, and a whole line of innovators faded, while Instagram has one billion users and was on track to earn more than $10.87 billion in advertising revenue in 2019.

Instagram started out just like those companies: an overnight hit, took investment, then gave no thought to making money. By 2011, Instagram's whole team came to just 13 people. Tumblr was the same — in its employees' words, its founder saw making money as "a boring thing." Snapchat lived on fundraising until its third year, when it ran its first ad.

The difference was that Instagram got sold to Zuckerberg. He may be one of the best businessmen among entrepreneurs.

As Ben Thompson put it: the road from product to company is a long one; Instagram's two founders, Kevin Systrom and Mike Krieger, only made it a product, while Zuckerberg made it a company.

Among all the new social apps that rose over these years, the only one to truly succeed commercially is Facebook.

The film The Social Network cast Zuckerberg as the stock founder, calculating that $1 million in advertising wasn't cool and that he should wait until the company was bigger and could earn $1 billion. But in reality, Facebook started running ads two months after launch, cleared $1 million in revenue that first year, and from the second year on, ad revenue grew faster than users did.

Had Instagram not accepted Facebook's acquisition, Kevin Systrom and Mike Krieger would have had to build their own sales team, collect users' location and preference data, open up ad slots, court advertisers, keep up those relationships, and slowly pile up ad fees to cover the company's costs. Instead they did none of it: they signed the contract, took the check, and plugged into Facebook's ad system.

In the end, because Facebook's demands never stopped — copy competitors' products, harvest users' private data, pile on the ads — the founders of both Instagram and WhatsApp left over the past year. But these products themselves have become immovable, and they've made Facebook immovable too.

Beer can have its craft brews; music can have its independent labels. Even in a government-controlled industry, you rarely see a monopoly as complete as these two companies'. But online social is a place where you're either the biggest or you fade — even companies like Snapchat, Twitter, and Weibo, with hundreds of millions of users, aren't enough.

The scale effect of social is one big reason, but just as important is the reality that the whole internet runs on advertising.

Making money from advertising is itself a huge barrier: only the biggest, most crowded places can earn enough through ads to sustain growth. (Gaming is just a route Tencent found to monetize its own ad inventory.)

In the end, the innovators couldn't stay, and social on the phone froze into two companies. They aren't only good at running a business and good at building products; they also know how to sell privacy and shut out their rivals.

Cover image / MySign AG on Visual Hunt / CC BY

Banner image / markchadwickart on Visualhunt.com / CC BY-NC-ND

Illustration / Feng Xiuxia