Talko

» see talko.com as it lived, on the wayback machine

~ the last words ~

ray ozzie spent most of the early 2010s building talko. he had left microsoft in 2010 as the chief software architect who briefly held bill gates's old title, and he had spent the prior thirty years building lotus notes and groove networks before that. talko was incorporated in january 2012. the public ios launch came on september 23, 2014, almost three years after the company started. the pitch was that voice would be the next great work-communication medium, and the existing tools, namely conference-call lines and threaded chat, were both wrong shapes for the actual problem.

the app worked by letting you press and hold to record voice into a thread shared with your work team. the thread persisted, so you could come back later and listen, jump to bookmarks other people had set, tag passages, and add text replies that would show up alongside the voice clips. the design language borrowed from the platform system that ozzie had worked on as a student in the 1970s. the name itself was an homage to talkomatic, an early group chat program ozzie had used on plato. it was a serious product with real engineering behind it. it also shipped late.

by 2015 the workplace-collab market was decisively moving toward text. slack had launched publicly in february 2014 and was already the default new-team tool for any startup. text was fast to scan, asynchronous in a way that voice was not, and produced searchable artifacts. voice messages did not. talko's pitch that voice would feel more human and less time-pressured was true in a way that did not matter, because the people choosing tools for their teams were choosing for searchability and scan-speed, not for warmth.

microsoft acquired talko on december 21, 2015. ozzie did not return to the company. the engineering team joined skype with the goal of porting talko's recording and bookmark features into skype and skype for business. the standalone app was shut down on march 16, 2016, less than three months after the acquisition closed. nothing recognizable from talko's product ever showed up in skype, which was already on a slow decline of its own. the closest living analog is voice messages in teams, which arrived years later through a different code lineage.

i was 31 when talko shipped and working on engineering teams that lived inside email plus an occasional skype call. talko's pitch never really applied to me. the workflow it was designed around, a globally distributed team of knowledge workers who needed to leave each other contextual voice notes across timezones, was a san francisco startup problem more than an ankara engineering problem. teams here resolved coordination questions in the office or over a phone call, and async-voice as a category never landed in the turkish workplace either.

~ the record ~

~ what ozzie was trying to build ~

the underlying thesis was that voice carried tone and nuance that text could not, and that tooling could close the gap between speakers by making voice rewindable and bookmarkable. that thesis is actually correct on the merits. by 2026 voice messages on whatsapp and telegram are routine for personal use, and discord and slack have both shipped voice-clip features. what talko got wrong was the bet on enterprise as the first beachhead. enterprise wanted text logs and searchability before it wanted warmth. the consumer market would have wanted talko more, but ozzie's network and instincts pointed him at the enterprise side, where the lotus notes and groove pedigree mattered. that pedigree did not translate to the 2014 product landscape.

~ the microsoft handoff ~

microsoft framed the acquisition as a skype tech-and-talent injection. the cleaner read is that it was an acquihire designed to bring ozzie-adjacent engineers back inside the building, with the side benefit of shutting down a competing-shape app before it had time to find a niche. ozzie himself stayed independent. by 2016 he was working on an unrelated iot project. the talko team got absorbed into the skype org chart at exactly the moment skype itself was losing internal priority to teams, which would absorb most of microsoft's workplace-communication ambitions by 2017. so the talko engineers landed inside a unit that was about to become unimportant. the recording and bookmark features they had built mostly stayed shelved.

~ the funny part ~

ozzie had been here before. groove networks shipped a careful peer-to-peer collaboration product in the late 1990s that microsoft bought in 2005 for $120 million. groove's tech got folded into sharepoint and onedrive and onenote in pieces, and ozzie became chief software architect under gates. talko followed the same arc thirteen years later with a smaller buyout and no architect job at the end. for ozzie personally talko is the second time he built a category-correct product that microsoft bought to defang rather than to ship. the part that is funny rather than sad is that he kept doing this voluntarily.

@}-->--

              )
             (_)
            .-'-.
            |   |
            |   |
            |   |
            |   |
          __|   |__   .-.
       .-'  |   |  `-:   :
      :     `---'     :-'
       `-._       _.-'
           '""""""

« previous  ·  random  ·  next »