Dead End: Google Reader

Abstract

Google Reader launched on October 7, 2005 as a web application for reading RSS and Atom feeds. It became the default way a generation of writers, developers, and news obsessives followed the web: subscribe to a site’s feed, and every new post arrived in one inbox-like list, sorted, searchable, and synced across devices. On March 13, 2013, Google announced it would shut Reader down on July 1, citing “declining” usage. A Change.org petition passed 100,000 signatures within days. Feedly took 500,000 refugees in the first 48 hours. The product had maybe tens of millions of users, small by Google’s standards and large by everyone else’s, and Google closed it anyway. The shutdown did two things at once: it dispersed the single largest audience RSS ever had, and it taught the web that a free Google service you built your habits around could vanish on ninety days’ notice. That second lesson outlived the first.

What Reader Was

RSS solved a specific problem. Instead of visiting twenty websites to check for new posts, a reader subscribed to each site’s feed, an XML file the site updated when it published, and a feed reader polled those files and collected the new items. The format had been through a decade of standards fights by the time Reader arrived (see Dave Winer and RSS). What it lacked was a mainstream client that ordinary people would actually use.

Google Reader supplied it. The project was built by Google engineer Chris Wetherell and a small team, and launched publicly on October 7, 2005. It stored subscriptions server-side, so the same read/unread state followed a user from office desktop to home laptop to phone. It imported and exported subscription lists as OPML, so switching in was easy. It added starred items, keyword search across everything a user had ever read, and folder labels. A mobile version arrived on May 18, 2006, an iPhone web app on May 12, 2008, and an Android app in December 2010.

For its core audience the appeal was speed. Keyboard shortcuts moved through hundreds of items a minute. Power users followed thousands of feeds and cleared them daily. Bloggers watched their subscriber counts in Reader as a proxy for reach, because Reader reported how many of its users subscribed to a given feed, a number that functioned as an informal circulation figure for the independent web.

The Sharing Layer

Reader was also, quietly, a social network. A sharing feature added in 2006 let users republish items to a public page and follow other people’s shared items. In December 2007 Google wired this into Google Talk, automatically sharing items with a user’s chat contacts, a move criticized at the time because there was no clear opt-out and it exposed reading habits to people users had never meant to share with.

Despite the rocky start, the sharing layer grew into a genuine community. Small networks of people followed each other’s shared items and comments, a curated feed of feeds. In some places it mattered more than convenience: users in Iran relied on Reader’s sharing to pass around news that the state censored, because the traffic looked like ordinary Google traffic and was hard to block selectively.

In October 2011 Google removed the native sharing features and replaced them with a +1 button tied to Google+, the social network Google was then trying to make happen. The existing follower networks were dismantled in the process. For the people who had built them, this was the first shutdown: the product still ran, but the part they valued was gone, sacrificed to prop up a different product. The reaction was angry enough that it became a preview of the reaction to come.

The Shutdown

On March 13, 2013, Google published a blog post titled “A second spring of cleaning.” Buried in a list of eight products being retired was Google Reader. The Reader team’s own farewell post, published the same day, gave the reason: “usage of Google Reader has declined, and as a company we’re pouring all of our energy into fewer products.” Users had until July 1 to export their data through Google Takeout and find somewhere else to go.

The declining-usage claim was hard to check and widely disbelieved. Reader had never carried ads and was never meant to make money directly; it was a feature that made the wider Google ecosystem stickier. By 2013 that logic had reversed. Google wanted attention flowing through Google+, and a beloved, ad-free, standards-based reader that sent users off to the open web was working against the strategy. Chris Wetherell, by then long gone from Google, told journalist Om Malik that executives had doubted the project from the start (the team kept a sign reading “days since cancellation”) and that it reached the market through sheer perseverance. When sharing was replaced with +1 in 2011, he said, “it was clear that this day was going to come.”

The user response was immediate and larger than Google seemed to expect. A Change.org petition to keep Reader alive passed 100,000 signatures. Competing services absorbed the exodus: Feedly reported 500,000 new users in the first 48 hours and 3 million within two weeks; the one-person NewsBlur watched its subscriber base jump from around 1,500 to 60,000 and buckled under the load; The Old Reader and, later, Digg Reader launched or scaled to catch the fallout. RSS reading survived. It just scattered across a dozen smaller services, none with Reader’s reach.

Why the Trust Broke

Google had killed products before. Google Wave was discontinued in 2010, Google Buzz in 2011. Reader was different because of what it was: not a failed experiment that never found users, but a working tool with a devoted audience that had woven it into a daily routine and, in the case of bloggers and publishers, into their livelihoods. Killing a flop is housekeeping. Killing something people love and depend on, for reasons of internal strategy, reads as a warning.

The phrase “Google kills products” hardened into a meme after Reader. The community site Killed by Google later catalogued the pattern; by then it listed dozens of discontinued services. The reasoning it encouraged was simple and durable: if Google could close Reader, it could close anything, so building a business or a habit on a free Google service was a risk. That skepticism followed Google into every later launch. When Google+ itself shut down in 2019, and when Stadia and others followed, the reaction always circled back to Reader as the moment the goodwill cracked.

The damage to RSS was structural. Reader had been the format’s center of gravity: the client most sites optimized for, the subscriber count publishers watched, the low-friction on-ramp that made following feeds normal for non-technical users. With that center gone, RSS did not die, but it stopped being something a mainstream user was likely to encounter. Feed reading receded into a tool for enthusiasts while the audience it had served drifted to algorithmic timelines on Facebook and Twitter, where a company decided what a user saw. In 2022 Techdirt called Reader’s closure “one of the defining moments in the shift from a more distributed, independent web to one that is controlled by a few large companies.”

What It Left Behind

The exodus seeded a small, stubborn ecosystem. Feedly, Inoreader, NewsBlur, and The Old Reader still run, most of them years after the panic that created their customer base. Feed reading remains a niche that refuses to disappear, kept alive by podcasts (which are RSS underneath), by newsletter fatigue, and by people who want a reading list no algorithm curates.

The deeper legacy is a habit of suspicion. After Reader, treating a free service as permanent came to look naive, and “will this be around in five years?” became a reasonable question to ask of anything Google shipped. Google built one of the best-loved tools of the 2000s web, and its most lasting effect was to teach a generation not to trust the company that made it.