Who killed Google Reader?

Who killed Google Reader?

Illustration by Hugo Herrera for The Verge

Ten years after its untimely death, the team that built the much-beloved feed reader reflects on what went wrong and what could have been.

There was a sign in the Google Reader team’s workspace at the company’s headquarters in Mountain View, California. “Days Since Cancellation,” it read, with a number below: zero. It was always zero.

This was in 2006 or so, back when Google Reader was still growing. Back when it still existed at all. Google’s feed-reading tool offered a powerful way to curate and read the internet and was beloved by its users. Reader launched in 2005, right as the blogging era went mainstream; it made a suddenly huge and sprawling web feel small and accessible and helped a generation of news obsessives and super-commenters feel like they weren’t missing anything. It wasn’t Google’s most popular app, not by a long shot, but it was one of its most beloved.

Within the company, though, Reader’s future always felt precarious. “It felt so incongruent,” says Dolapo Falola, a former engineer on the Reader team. “Literally, it felt like the entire time I was on the project, various people were trying to kill it.”

Of course, Google did kill it. (Google didn’t respond to a request for comment on this story.) Reader’s impending shutdown was announced in March of 2013, and the app went officially offline on July 1st of that year. “While the product has a loyal following, over the years usage has declined,” Google SVP Urs Hölzle wrote in a blog post announcing the shutdown.

Google tried its best to bury the announcement: it made it the fifth bullet in a series of otherwise mundane updates and published the blog post on the same day Pope Francis was elected to head the Catholic Church. Internally, says Mihai Parparita, who was one of Reader’s last engineers and caretakers, “they were like, ‘Okay, the Pope will be the big story of the day. It’ll be fine.’ But as it turns out, the people who care about Reader don’t really care about the Pope.” That loyal following Hölzle spoke of was irate over losing their favorite web consumption tool.

Google’s bad reputation for killing and abandoning products started with Reader and has only gotten worse over time. But the real tragedy of Reader was that it had all the signs of being something big, and Google just couldn’t see it. Desperate to play catch-up to Facebook and Twitter, the company shut down one of its most prescient projects; you can see in Reader shades of everything from Twitter to the newsletter boom to the rising social web. To executives, Google Reader may have seemed like a humble feed aggregator built on boring technology. But for users, it was a way of organizing the internet, for making sense of the web, for collecting all the things you care about no matter its location or type, and helping you make the most of it.

A decade later, the people who worked on Reader still look back fondly on the project. It was a small group that built the app not because it was a flashy product or a savvy career move — it was decidedly neither — but because they loved trying to find better ways to curate and share the web. They fought through corporate politics and endless red tape just to make the thing they wanted to use. They found a way to make the web better, and all they wanted to do was keep it alive.

Photo by Chris Wetherell

From left to right: Ben Darnell, Chris Wetherell, and Laurence Gonsalves, three of the early members of the Reader team.

“I think I built a thing”

“This is going to be the driest story ever,” says Chris Wetherell, when I ask him to describe the beginning of Google Reader. Wetherell wasn’t the first person at Google to ever dream of a better way to read the internet, but he’s the one everyone credits with starting what became Reader. “Okay, here goes: a raging battle between feed formats,” he says when I push. “Does that sound interesting?”

Here’s the short version: One of the most important ways that information moves around the internet is via feeds, which automatically grab a webpage’s most important content and make it available. Feeds are what make podcasts work across apps, and how content shows up in everything from Flipboard to Facebook. In the early aughts, there were basically two ways to build a feed. One was RSS, which stands for Really Simple Syndication and has been around approximately forever. The other was called Atom, a newer standard that aimed to fix a lot of the things that were outdated and broken with RSS.

In late 2004, Jason Shellen, a product manager working on Atom projects at Google, called up Wetherell, a former colleague on the Blogger team, and asked him if he could hack together some kind of Atom-based app. “Is there any way you could write just a little thing that would parse Atom, just to show how it works?” Shellen asked. All he really needed was a tech demo, something he could show potential partners to explain how Atom worked.

Wetherell stayed up late one night building a simple app that converted a bunch of websites’ RSS feeds to Atom and displayed those feeds in a Javascript-based browser app so you could click around and read. “And then I tried to make it a pleasant arrangement,” Wetherell says. He called it Fusion. It wasn’t much to look at, but it was fast and worked in a web browser.

Photo by Chris Wetherell

Wetherell’s first prototype didn’t look like much, but it felt like nothing before.

And then the strangest thing happened: as soon as he’d finished the Fusion app, Wetherell started using it to actually read stuff from the sites whose feeds he’d grabbed. He turned to his partner that night and said: “I think I built a thing.” Wetherell sent the prototype to Shellen, who also immediately saw its potential.

In 2004, most people weren’t viewing the internet through a bunch of social networks and algorithmic feeds. Facebook and Twitter were barely blips on the radar. At that point, most people experienced the internet by typing in URLs and going to websites. A few tools like NetNewsWire and Bloglines had cropped up to make it easier to subscribe to lots of sites in one place, but these RSS readers were mostly tools for nerds. Most users were stuck managing bookmarks and browser windows and furiously refreshing their favorite sites just to see what was new. Wetherell’s prototype wasn’t complicated like NetNewsWire, it didn’t crash like Bloglines, and the Javascript interface felt fast and smooth. It immediately felt like a better way to keep up with the web.

Wetherell and Shellen started imagining all the different kinds of feeds this tool could store. He thought it might bring in photo streams from Flickr, videos from YouTube and Google Videos, even podcasts from around the web. Shellen, who had come to Google as part of its Blogger acquisition, saw the possibility for a social network, a single place to follow all your friends’ blogs. “Of course, it was just a hacky list of feeds,” Wetherell says, but there was something about the speed with which you could flip through articles and headlines, the information density, the simplicity of the reading experience, that just worked.

Ultimately, Wetherell ended up spending some of his 20 percent time — Google’s famous policy of letting employees work on just about whatever they wanted, which ironically died about the same time Reader did — building Fusion into a more complete feed-reading product. It handled RSS, Atom, and more. After a while, he wound up showing it to the folks building iGoogle, the company’s recently launched web-homepage product. (iGoogle has since been killed, of course.)

Image by Chris Wetherell

As the Fusion prototypes got more polish, they started to look more like Reader.

In his presentation, Wetherell shared a much bigger, grander ambition for Fusion than an article-reading service. He and Shellen had been talking about the fact that a feed could be, well, anything. Wetherell kept using the word “polymorphic,” a common term in programming circles that refers to a single thing having many forms.

“I drew a big circle on the whiteboard,” he recalls. “And I said, ‘This is information.’ And then I drew spokes off of it, saying, ‘These are videos. This is news. This is this and that.’” He told the iGoogle team that the future of information might be to turn everything into a feed and build a way to aggregate those feeds.

Fusion was meant to be a social network based on content, on curation, on discussion

The pitch sounded good, and they got permission to keep working on it. Fusion wasn’t exactly made an official project or staffed like one, but it was at least allowed to continue to exist. Wetherell and Shellen recruited other people working on similar projects in their 20 percent time, and Shellen wrote an official product spec document outlining Fusion’s ambitions. The vision, he wrote, was to “become the world’s best collaborative and intelligent web content delivery service.” It promised to “build a robust web service and best-of-breed user interface for viewing subscriptions” and to produce an API that would let other apps tap into the same underlying data.

In other words, Fusion was meant to be a social network. One based on content, on curation, on discussion. In retrospect, what Shellen and Wetherell proposed sounds more like Twitter or Instagram than an RSS reader. “We were trying to avoid saying ‘feed reader,’” Shellen says, “or reading at all. Because I think we built a social product.”

That was the idea, anyway.

Google goes social

In October of 2005, Shellen announced Fusion to the world at the Web 2.0 Conference in San Francisco. Only he wasn’t allowed to call it Fusion. The team had been forced to change the name at the last minute, after Marissa Mayer — at that point, the Google executive in charge of all of the company’s consumer web services — said she wanted the name for another product and demanded the team pick another one. (That product never launched, and nobody I spoke to could even remember what it was. Mayer also didn’t respond to a request for comment.)

The team had brainstormed dozens of other names: Reactor, Transmogrifier, and a whiteboard full of others. Down near the bottom of the list: Reader, “a name none of us liked,” Wetherell says, “because it does many other things, but… it’s fine.” But somehow, that became the choice.

Shellen in particular still rues losing the fight over the name. Even now, he bristles thinking about the fight and the fact that Google Reader is known as “an RSS reader” and not the ultra-versatile information machine it could have become. Names matter, and Reader told everyone that it was for reading when it could have been for so much more. “If Google made the iPod,” he says, “they would have called it the Google Hardware MP3 Player For Music, you know?”

So Fusion launched, as Google Reader, and immediately crashed spectacularly. The site simply couldn’t keep up with the traffic on the first day. Most of those early visitors to reader.google.com never came back, either. Even once the Reader team stabilized the infrastructure, lots of users hated the product; it had a lot of clever UI tricks but just didn’t work for too many users. “People don’t remember this,” Wetherell says, “but it bombed. It was terrible. We were accused by someone of hurting the share price of Google because it bombed so hard.”

It wasn’t until the team launched a redesign in 2006 that added infinite scrolling, unread counts, and some better management tools for heavy readers that Reader took off. Another newish Google product, Gmail, had far more users, but the engagement with Reader was off the charts. “People would spend, I don’t know, five minutes a day on iGoogle,” Parparita says, “and like an hour a day in Reader.” The team hadn’t been pushed to worry about monetization or user growth, but they felt like they were on the right track.

Reader appealed primarily to information junkies, who wanted a quick way to keep up with all their favorite publications and blogs. (It turned out there were two types of Reader users: the completionists, who go through every unread item they have, and the folks who just scroll around until they find something. Both sides think the other is bonkers.) The team struggled to find ways to bring in more casual users, some of whom were put off by the idea of finding sites to subscribe to and others who simply didn’t care about reading hundreds of articles a day.

One feature took off immediately, for power users and casual readers alike: a simple sharing system that let users subscribe to see someone else’s starred items or share their collection of subscriptions with other people. The Reader team eventually built comments, a Share With Note feature, and more. All this now seems trite and obvious, of course, but at the time, a built-in way to see what your friends liked was novel and powerful. Reader was prescient.

Reader was always a power-user tool at heart, and it appealed to people with a lot to read.

At its peak, Reader had just north of 30 million users, many of them using it every day. That’s a big number — by almost any scale other than Google’s. Google scale projects are about hundreds of millions and billions of users, and executives always seemed to regard Reader as a rounding error. Internally, lots of workers used and loved it, but the company’s leadership began to wonder whether Reader was ever going to hit Google scale. Almost nothing ever hits Google scale, which is why Google kills almost everything.

The bigger problem seemed to be that Mayer didn’t like it: Shellen says she told him at one point that he was wasting his engineers’ careers working on Reader. The team had trouble getting face time in product reviews, and asking for additional resources or funding was a waste of time. Google co-founder Larry Page had been a fan of the app — Jenna Bilotta, a designer on the team, remembers he had this very specific idea about using Reader to research windmill-generated energy — but a few years later, Shellen recalls Reader appearing on Page’s list of Google’s worst 100 projects.

Google’s executives always seemed to think Reader was a feature, not a product. In meeting after meeting, they’d ask why Reader wasn’t just a tab in the Gmail app. When a team decided to build a new email client called Inbox, with promises of collecting all your important communication and information in one place, executives thought maybe Reader should be part of that. (Inbox was eventually killed, too.)

Every so often, a faction of the Reader team was called into a meeting and asked to justify the product’s ongoing existence. It didn’t require many resources, which was helpful; the team only ever got as big as about a dozen people, many of them on loan from other teams at the company. On the other hand, Reader wasn’t a roaring Google scale success, nor did it have a powerful executive championing its existence. It seemed the company got more tired of this side project all the time. Falola still remembers one particularly telling interaction: “We were having some back and forth with some VP at the time, making our petition for why you should keep Reader around, and I remember that VP responding with, ‘Don’t confuse this for a conversation between peers.’”

Threatened by the rise of social networks — namely Facebook and its quickly encroaching seizure of the online ad market — Google became desperate to build its own. It tried to build a social graph called Google Friend Connect, which went nowhere. It decided to build a network around email contacts, where the company already had a head start because of Gmail, but that didn’t make any sense. So the company’s big swing became Google Buzz, an app that tried to combine messaging, social networking, and blogging into one thing. That launched in 2010 and was killed in 2011.

For a while, the Reader team managed to stay alive by promising to be the guinea pig for Google’s other social ideas. It tried the Gmail contacts thing; Parparita remembers that as “the year Reader ruined Christmas” because the feature launched in December and suddenly everyone’s mom and landlord and Craigslist acquaintance could see all the articles they’d starred. (The team scrambled to build sharing management tools quickly after that.) The Reader engineers worked with the Buzz team, the iGoogle team… anyone who needed help.

The tide turned when Google decided not just to build a social product but to fundamentally re-architect the company’s apps around social. Two executives, Vic Gundotra and Bradley Horowitz, started a new project codenamed “Emerald Sea” with plans to build sharing and friend-based recommendations into just about every Google app. It would come to be known as Google Plus, the company’s most direct shot at a Facebook-like product, and Gundotra and Horowitz amassed an empire within the company. “We’re transforming Google itself into a social destination at a level and scale that we’ve never attempted — orders of magnitude more investment, in terms of people, than any previous project,” Gundotra told Wired in 2011. He wasn’t exaggerating.

“As far as I could tell, nobody ever won against them,” Parparita says. “They just got their own way.” There was plenty of opposition to the project, including from the Reader team, but it didn’t matter. The Emerald Sea team worked in a special building, only accessible to a few employees; the secrecy was just one more signal to everyone that this was Google’s top priority.

“As far as I could tell, nobody ever won against [Google Plus]. They just got their own way.”

Gundotra and Horowitz also seemed to pluck any employee they wanted. And they wanted a number of Reader employees, who were some of Google’s most well regarded. “We assembled the Beatles,” says Wetherell, and Shellen calls the team a “Murderer’s Row.” Both singled out Paraprita as one of Google’s best engineers, along with Ben Darnell, a back-end whiz who built much of the product’s underlying infrastructure. Many of these engineers had started working on Reader as a side project, simply because they loved the app. Some had done stints full-time and then gone on to other projects. Now it felt like everyone was being pulled into Plus — and many of them chose to leave the company instead.

And in its effort to build a splashy new social platform, the Reader team felt Google was missing the burgeoning one right under its nose. Reader was probably never going to become the world-conquering beast Facebook eventually became, but the team felt it had figured out some things about how people actually want to connect. “There were people that met on Google Reader that got married,” Bilotta says. “There are whole communities that met on Google Reader that meet up — they fly to meet each other! It was crazy. We didn’t anticipate this being that sticky.” The team was plotting new ways for users to discover content, new tools for sharing, and more. Bilotta urged executives to see the potential: “They could have taken the resources that were allocated for Google Plus, invested them in Reader, and turned Reader into the amazing social network that it was starting to be.”

By early 2011, with the team severely diminished, Reader had been officially put into “maintenance mode,” which meant that an engineer — Parparita, mostly — would fix anything spectacularly broken but the product was otherwise to be left alone. Reader was integrated into Google Plus, sort of, before Plus began its inexorable decline. Despite Google practically force-feeding its social network to hundreds of millions of people, users rebelled against Google’s real-name policy, resented its spam problem, and ultimately could never figure out what Plus could do that Facebook or Twitter couldn’t. “The engagement was so low,” Bilotta says, “that basically within eight months, they realized it wasn’t going to be a product.”

The damage was done for Reader, though. Its core team was gone, its product had withered, and by the end of 2012, even Parparita had left Google. Hardly anyone on the team was surprised when Google announced a few months later that Reader was shutting down for good.

The alternate Reader universe

It’s been a decade since Reader went offline, and a number of the folks who helped build it still ask themselves questions about it. What if they’d focused on growth or revenue and really tried to get to Google scale? What if they’d pushed harder to support more media types, so it had more quickly become the reader / photo viewer / YouTube portal / podcast app they’d imagined? What if they’d convinced Mayer and the other executives that Reader wasn’t a threat to Google’s social plans, but actually could be Google’s social plans? What if it hadn’t been called Reader and hadn’t been pitched as a power-user RSS feed aggregator?

And, of course, there’s the biggest question: what if they’d tried to build Reader outside of Google? It had millions of devoted users, a top-notch team, and big plans. “At that time, outside of Google, VCs would have been throwing money at us left and right,” Wetherell says. Inside Google, it could never compete. Outside Google, there would have been no politics, no crushing weight of constant impending doom. If Google had been driven by anything other than sheer scale, Reader might have gotten to Google scale after all.

The biggest question: what if they’d tried to build Reader outside of Google?

But Reader was also very much a product of Google’s infrastructure. Outside Google, there wouldn’t have been access to the company’s worldwide network of data centers, web crawlers, and excellent engineers. Reader existed and worked because of Google’s search stack, because of the work done by Blogger and Feedburner and others, and most of all, the work done by dozens of Google employees with 20 percent of their time to spare and some ideas about how to make Reader better. Sure, Google killed Reader. But nearly everyone I spoke to agreed that without Google, Reader could never have been as good as it was.

Over the years, people have approached Bilotta, Falola, and a few of the other ex-Reader team members about building something in the same vein. Shellen and Wetherell ended up co-founding Brizzly, a social platform based on a lot of the ideas in Reader. Kevin Systrom, once a product marketing manager on the Reader team, went on to found Instagram and, more recently, Artifact, two platforms with big ideas about information consumption that clearly learned from what went wrong at Reader.

For a while, the internet got away from what Google Reader was trying to build: everything moved into walled gardens and algorithmic feeds, governed by Facebook and Twitter and TikTok and others. But now, as that era ends and a new moment on the web is starting to take hold through Mastodon, Bluesky, and others, the things Reader wanted to be are beginning to come back. There are new ideas about how to consume lots of information; there’s a push toward content-centric networks rather than organizing everything around people. Most of all, users seem to want more control: more control over what they see, more knowledge about why they’re seeing it, and more ability to see the stuff they care about and get rid of the rest.

Google killed Reader before it had the chance to reach its full potential. But the folks who built it saw what it could be and still think it’s what the world needs. It was never just an RSS reader. “If they had invested in it,” says Bilotta, “if they had taken all those millions of dollars they used to build Google Plus and threw them into Reader, I think things would be quite different right now.”

Then she thinks about that for a second. “Maybe we still would have fallen into optimizing for the algorithm,” she allows. Then she thinks again. “But I don’t think so.”