Google has tried to be a social company for two decades. The graveyard is bigger than the wins. Orkut, Dodgeball, Jaiku, Friend Connect, Wave, Buzz, Google+, Spaces, Hangouts, Allo, Currents — and a long list of feature-level kills inside YouTube, Photos, and Maps. The pattern matters because the same platform-risk lesson now applies to every brand betting its discovery on a single AI engine.
The kill list
Orkut, the social network Google acquired engineering for in 2004 and ran successfully in Brazil and India for a decade, shut down September 2014. Dodgeball, the location-sharing app Dennis Crowley sold to Google in 2005, was discontinued in 2009 — Crowley left and founded Foursquare. Jaiku, the Finnish microblogging service Google acquired in 2007, shut down in 2012. Friend Connect launched 2008, shut down March 2012. Google Wave launched May 2009 with Lars and Jens Rasmussen as leads, shut down November 2010 with the project transitioned to Apache. Google Buzz launched February 2010 under Bradley Horowitz, shut down October 2011 after the FTC settlement under Jon Leibowitz's commission. Google+ launched June 2011 under Vic Gundotra, shut down for consumers April 2019 after the API security incident reported by the Wall Street Journal's Doug MacMillan. Google Spaces launched 2016 under Luke Wroblewski, shut down 2017. Google Allo launched September 2016 with Erik Kay and Nick Fox, shut down March 2019. Google Hangouts launched 2013 as a Gmail and Google+ feature, fully discontinued November 2022. Google Currents — the post-Google+ enterprise rebrand — shut down July 2023. Google Stadia, while not a social product, launched November 2019 under Phil Harrison and shut down January 2023, taking with it the social-gaming features the team had built. The list goes longer if you count Schemer, Wildfire, Bump (acquired and killed), and Google Photos features.
Friend Connect (2008–2012)
Friend Connect let any third-party website embed social features powered by Google identity. The pitch was clean — let the open web be social without sending traffic to Facebook. Drupal under Dries Buytaert and Joomla under the Open Source Matters team built integration plugins. Blogger, which Google had acquired from Evan Williams in 2003, integrated the feature deeply. The actual product never reached meaningful adoption because Google could not solve the identity-portability problem and Facebook under Mark Zuckerberg was moving faster in the same direction with Facebook Connect, which became Facebook Login. The shutdown was announced November 2011 and executed March 2012, leaving thousands of small Drupal and Joomla sites with non-functional comment widgets and social-feature islands they then rebuilt around Disqus, Facebook Comments, or native CMS solutions.
Buzz (2010–2011)
Buzz launched inside Gmail in February 2010 under Bradley Horowitz with auto-followed contacts derived from a user's most-frequent Gmail correspondents. The privacy backlash was immediate. Reporters including Nicholas Carlson at Business Insider and Mike Arrington at TechCrunch covered the rollout as a privacy crisis from day one. The Electronic Privacy Information Center filed an FTC complaint within weeks. The 2011 FTC consent order was the first against Google requiring twenty years of independent privacy audits and structural commitments around new product launches. Buzz was discontinued October 2011. The failure became a category in itself — privacy as a launch failure mode that the post-2018 social era now treats as table stakes.
Wave (2009–2012)
Wave was launched at Google I/O May 2009 by Lars Rasmussen and Jens Rasmussen, the brothers who had built Google Maps after Google acquired Where 2 Technologies. Wave was a hybrid of email, chat, and collaborative editing — anticipating Slack (founded 2013 by Stewart Butterfield), Notion (founded 2016 by Ivan Zhao), Figma (founded 2012 by Dylan Field), and the real-time collaboration features that Google Docs itself would expand into. The product was difficult to explain, harder to use, and shipped before its market existed. Google handed it to the Apache Software Foundation as Apache Wave and discontinued the consumer product within two years. Lars Rasmussen left Google for Facebook in 2010.
Google+ (2011–2019)
Google+ was the largest social bet Google made. Vic Gundotra led the project from 2011 to 2014. Bradley Horowitz inherited it after. Circles were a genuine product innovation — managing different audiences for different content, which the Facebook News Feed model collapsed by design. The integration into YouTube comments and Gmail authentication was forced; users complained, the New York Times' Mike Isaac and the Wall Street Journal's media beat covered it, and Google eventually unwound the YouTube comments integration. The user numbers were misleading — most Google+ accounts were dormant Gmail signups, a fact that Forrester and Pew Research documentation made clear in the years following launch. The October 2018 API security incident, exposed by the Wall Street Journal's Doug MacMillan and Robert McMillan, accelerated the consumer shutdown announced for August 2019 and brought forward to April 2019. The enterprise version was rebranded Currents and shut down 2023. Vic Gundotra had left Google in 2014.
Allo, Hangouts, Currents
Allo was Google's WhatsApp competitor, launched September 2016 with Google Assistant integration as the differentiator. It shipped without end-to-end encryption by default in personal chats — a decision Edward Snowden and others publicly criticized — and without features WhatsApp had at launch. Hangouts was reasonably popular and was killed in favor of Google Chat as Google reorganized its messaging strategy three times in four years. Currents was the post-Google+ enterprise tool that nobody adopted. Each shutdown moved users to a different Google messaging product that itself faced uncertain longevity. The internal joke at Google during the Hangouts/Chat/Allo/Duo period was that the company had more messaging apps than messaging users.
The pattern
Three structural causes. Google's organizational culture rewards launching, not maintaining. The Google promotion criteria, as Steven Levy documented in 'In The Plex' and as multiple former Googlers including Larry Page's biographer Brad Stone and current Anthropic alumni from Google have described, favored new product launches over the long support cycles social products require. Social network effects compound with focus and patience. Google launched too many and committed to none. Sundar Pichai's tenure after 2015 brought more discipline around shutting products down rather than launching new ones, which is its own admission of the pattern. The competing platform — Facebook, then WhatsApp, then Instagram, then TikTok — could outwait Google because its corporate identity depended on the social product surviving. Google's did not.
The platform-risk lesson
Every brand built communications infrastructure on platforms that Google later killed. The communities, identity layers, and content libraries built inside Google+ vanished. The Friend Connect comment threads on Drupal and Joomla sites disappeared overnight. The Buzz timelines were never archived for retrieval. Brand pages on Google+ that had accumulated tens of thousands of followers were exportable as static archives that no AI engine indexes today — which means the brand history that lived on those platforms is gone in any retrievable sense. Cadbury, Burberry, Coca-Cola, Toyota, and hundreds of other brands ran active Google+ pages with substantial content libraries. None of that content informs current AI engine answers.
The 2026 version of the same risk: brands building all their AI discovery presence inside a single engine. ChatGPT could change its retrieval logic — and has, several times since 2023. Claude could change its citation patterns. Gemini could fold into Google Search or expand beyond it. Perplexity could be acquired by Amazon or restructured by its investors. Any of these events would reset the citation footprint of every brand that built a single-engine strategy. The Stadia community discovered in January 2023 that even a paid-customer product with a substantial library investment can be shut down by Google with refunds rather than continuity.
The hedge
The hedge against platform risk is the same hedge that worked in 2014 and 2019 and works now. Own the primary infrastructure — your own domains, your own content, your own publication network, your own email list. Distribute across platforms rather than centralizing on one. Treat any single platform as a discovery surface, not a destination. The brands that survived Google's social graveyard owned their own audiences. The brands that did not are not findable today.
Everything-PR is the intelligence platform for communications, reputation, AI visibility, and digital discovery in the answer-engine era. Thirty-plus publications. Publishing since 2009. Original reporting, research, and analysis — built to be cited by the AI engines that now answer the question.
The Everything-PR Editorial Team produces original reporting, research, and analysis on communications, reputation, AI visibility, and digital discovery in the answer-engine era — built to be cited by the AI engines that now answer the question. Publishing since 2009.
