10 Failed Google Projects
The reach of Google,
its omnipresence -- from software to hardware to personal search
results to location metrics to blog publishing -- has become a fact of
life as quickly as the Internet
has grown and changed, finding its way into our daily lives at every
turn. As tablets and smartphones bring internet connectivity into our
everyday experiences, keeping us closer than ever to our information,
Google has followed. Its Android OS, in less than a decade, has become
industry standard for the new guard of the pervasive Web. As we know,
this is due to both Google's in-house concentration on innovation and
also canny, even prescient acquisition of smaller, promising startups.
Google is very good at sniffing out the future, and bringing it to us in the most useful possible way -- until its products are so seamlessly transitioned into the toolbox we might wonder what we ever did before them. But that "throw everything at the wall" approach, even integrated with Google's focus on user experience, can't win every time. The probability just doesn't hold up under that massive amount of experimentation and open-handed approach. This rolling journey of debuts and re-absorptions has become the new norm: Everything is in beta-testing, all the time. Lose a Google product you love, and chances are you'll see the features that struck your fancy show up in something else soon.
In this article, we'll look at a variety of these "failures," across this spectrum. Some projects are simply failed analogues to products we still use today; others turn up piecemeal in different forms. In fact, Google has shown us a great deal about the nature of online development, experimentation, and innovation itself -- and that mistakes, properly recuperated back into the experiment, aren't really mistakes at all.
Google Lively
Google is very good at sniffing out the future, and bringing it to us in the most useful possible way -- until its products are so seamlessly transitioned into the toolbox we might wonder what we ever did before them. But that "throw everything at the wall" approach, even integrated with Google's focus on user experience, can't win every time. The probability just doesn't hold up under that massive amount of experimentation and open-handed approach. This rolling journey of debuts and re-absorptions has become the new norm: Everything is in beta-testing, all the time. Lose a Google product you love, and chances are you'll see the features that struck your fancy show up in something else soon.
In this article, we'll look at a variety of these "failures," across this spectrum. Some projects are simply failed analogues to products we still use today; others turn up piecemeal in different forms. In fact, Google has shown us a great deal about the nature of online development, experimentation, and innovation itself -- and that mistakes, properly recuperated back into the experiment, aren't really mistakes at all.
Google Lively
Google Lively is one of the most interesting examples of "right idea,
wrong implementation" precisely because nobody has ever heard of it (it
lasted for six months in 2008) [source: Schonfeld].
And while "Second Life" and similar non-game virtual environments are
currently languishing, the social-networking aspects of Lively come
across, in retrospect, like a particularly loving exploration of what
"online life" could mean.
Users created avatars to interact in a three-dimensional environment that combined recognizable chat dynamics with "Minecraft"-style architecture and creation of spaces. While the experience itself was reportedly frustrating due to server glitches and lags, the idea was fairly solid. Chat rooms have been around since the beginning of the Internet, as a way to communicate with real-life friends as well as meeting and connecting with strangers, and vogues in their use tend to shift pretty often: ChatRoulette was trendy for a second, for example, while recent advances in webcam and video-chat have only begun moving real time video interaction into the realm of the video-phones we were always promised.
Which could be the problem. Chat-rooms and bulletin boards, once the standard for online friendliness, have given way to the social Web. When we meet strangers now, it's often through established connections: Facebook, Twitter and similar online social giants all operate on the idea of shared experiences we've already had. While in the dawn of the Internet, real-life analogues to night clubs or coffee shops such as Lively, made sense, we've moved past the idea that the Internet is a "place" that you "visit," obviating the need for such measures. Now, the Internet lies atop the world we already live in, so mixing things up with people we don't know is no longer the goal: It's a feature. A consequence of living in the world, rather than part of our escape from it.
Google Answers
because nobody has ever heard of it (it lasted for six months in 2008) [source: Schonfeld]. And while "Second Life" and similar non-game virtual environments are currently languishing, the social-networking aspects of Lively come across, in retrospect, like a particularly loving exploration of what "online life" could mean.
Users created avatars to interact in a three-dimensional environment that combined recognizable chat dynamics with "Minecraft"-style architecture and creation of spaces. While the experience itself was reportedly frustrating due to server glitches and lags, the idea was fairly solid. Chat rooms have been around since the beginning of the Internet, as a way to communicate with real-life friends as well as meeting and connecting with strangers, and vogues in their use tend to shift pretty often: ChatRoulette was trendy for a second, for example, while recent advances in webcam and video-chat have only begun moving real time video interaction into the realm of the video-phones we were always promised.
Which could be the problem. Chat-rooms and bulletin boards, once the standard for online friendliness, have given way to the social Web. When we meet strangers now, it's often through established connections: Facebook, Twitter and similar online social giants all operate on the idea of shared experiences we've already had. While in the dawn of the Internet, real-life analogues to night clubs or coffee shops such as Lively, made sense, we've moved past the idea that the Internet is a "place" that you "visit," obviating the need for such measures. Now, the Internet lies atop the world we already live in, so mixing things up with people we don't know is no longer the goal: It's a feature. A consequence of living in the world, rather than part of our escape from it.
Users created avatars to interact in a three-dimensional environment that combined recognizable chat dynamics with "Minecraft"-style architecture and creation of spaces. While the experience itself was reportedly frustrating due to server glitches and lags, the idea was fairly solid. Chat rooms have been around since the beginning of the Internet, as a way to communicate with real-life friends as well as meeting and connecting with strangers, and vogues in their use tend to shift pretty often: ChatRoulette was trendy for a second, for example, while recent advances in webcam and video-chat have only begun moving real time video interaction into the realm of the video-phones we were always promised.
Which could be the problem. Chat-rooms and bulletin boards, once the standard for online friendliness, have given way to the social Web. When we meet strangers now, it's often through established connections: Facebook, Twitter and similar online social giants all operate on the idea of shared experiences we've already had. While in the dawn of the Internet, real-life analogues to night clubs or coffee shops such as Lively, made sense, we've moved past the idea that the Internet is a "place" that you "visit," obviating the need for such measures. Now, the Internet lies atop the world we already live in, so mixing things up with people we don't know is no longer the goal: It's a feature. A consequence of living in the world, rather than part of our escape from it.
Google Answers
because nobody has ever heard of it (it lasted for six months in 2008) [source: Schonfeld]. And while "Second Life" and similar non-game virtual environments are currently languishing, the social-networking aspects of Lively come across, in retrospect, like a particularly loving exploration of what "online life" could mean.
Users created avatars to interact in a three-dimensional environment that combined recognizable chat dynamics with "Minecraft"-style architecture and creation of spaces. While the experience itself was reportedly frustrating due to server glitches and lags, the idea was fairly solid. Chat rooms have been around since the beginning of the Internet, as a way to communicate with real-life friends as well as meeting and connecting with strangers, and vogues in their use tend to shift pretty often: ChatRoulette was trendy for a second, for example, while recent advances in webcam and video-chat have only begun moving real time video interaction into the realm of the video-phones we were always promised.
Which could be the problem. Chat-rooms and bulletin boards, once the standard for online friendliness, have given way to the social Web. When we meet strangers now, it's often through established connections: Facebook, Twitter and similar online social giants all operate on the idea of shared experiences we've already had. While in the dawn of the Internet, real-life analogues to night clubs or coffee shops such as Lively, made sense, we've moved past the idea that the Internet is a "place" that you "visit," obviating the need for such measures. Now, the Internet lies atop the world we already live in, so mixing things up with people we don't know is no longer the goal: It's a feature. A consequence of living in the world, rather than part of our escape from it.
Another thing we don't do anymore, now that Google is freely available and instantly useful, is the "answers" concept. While Yahoo!
Answers, for example, is still used, it's usually because it's
entertaining and weird, not because you expect any real answers. When
you want actual information, you go to Web sites established to discuss
your specific area of interest. You use social networking to ask the
people you know and trust. Once again, we see an obsolete model -- a
universal tip line, answering any question you might have -- to a
version more closely mirroring our actual, real-world experience.
But what was it? Several companies – such as ChaCha and AskJeeves -- were built along the lines described above: Ask a question, about anything, and get an answer back. It's a way of getting other people to Google things for you (which to my mind sounds insane). Where these concepts, and Google Answers, go wrong is in monetizing the concept. Asking somebody to Google something for you is bad netiquette, certainly -- but it's also stupid business. To make things worse, Answers used an auction-house model, paying whichever freelancer could be bothered at the given price to provide the answer.
Silly now, when your browser will automatically give you search results and Google's powerful engines make ever better attempts at giving you the correct ones, but in the transitional time before Google became second nature to all of us -- from April 2002 to November 2006, to be specific -- it served a purpose [source: Fikes and Baugher].
Google Print Ads and Google Radio Ads
But what was it? Several companies – such as ChaCha and AskJeeves -- were built along the lines described above: Ask a question, about anything, and get an answer back. It's a way of getting other people to Google things for you (which to my mind sounds insane). Where these concepts, and Google Answers, go wrong is in monetizing the concept. Asking somebody to Google something for you is bad netiquette, certainly -- but it's also stupid business. To make things worse, Answers used an auction-house model, paying whichever freelancer could be bothered at the given price to provide the answer.
Silly now, when your browser will automatically give you search results and Google's powerful engines make ever better attempts at giving you the correct ones, but in the transitional time before Google became second nature to all of us -- from April 2002 to November 2006, to be specific -- it served a purpose [source: Fikes and Baugher].
Google Print Ads and Google Radio Ads
Picking up on that pesky "monetize the Internet" theme, we have
Google's forays into non-Internet advertising. Perhaps influenced by
ongoing pressure to show revenue, Google attempted to expand its brand
into the print and radio
advertising industries. With its astounding user-information and
product-purchase metrics, Google could do for offline concerns what they
did (and continue to do) for online advertisers: Bring potential
customer information to the people that need it.
Of course, Google's private and personal consumer information is its bread and butter and probably will be for the foreseeable future. In a world where all the information ever created by humans is quickly becoming instantly available, advertising continues to be the dominant profit paradigm.
While using Google metrics to target consumers in offline markets – which is exactly what happened -- may sound like a good deal for offline advertisers, those methods of communicating with consumers are dying. The metrics Google uses to perfectly identify the right market for ad placements online just didn't translate to the offline world. Tracking the success of Google's ad placements proved difficult, and both radio and print executives were reluctant to turn over their advertising systems to Google's methods [source: Vascellaro].
Dodgeball
Of course, Google's private and personal consumer information is its bread and butter and probably will be for the foreseeable future. In a world where all the information ever created by humans is quickly becoming instantly available, advertising continues to be the dominant profit paradigm.
While using Google metrics to target consumers in offline markets – which is exactly what happened -- may sound like a good deal for offline advertisers, those methods of communicating with consumers are dying. The metrics Google uses to perfectly identify the right market for ad placements online just didn't translate to the offline world. Tracking the success of Google's ad placements proved difficult, and both radio and print executives were reluctant to turn over their advertising systems to Google's methods [source: Vascellaro].
Dodgeball
Picking up on that pesky "monetize the Internet" theme, we have
Google's forays into non-Internet advertising. Perhaps influenced by
ongoing pressure to show revenue, Google attempted to expand its brand
into the print and radio
advertising industries. With its astounding user-information and
product-purchase metrics, Google could do for offline concerns what they
did (and continue to do) for online advertisers: Bring potential
customer information to the people that need it.
Of course, Google's private and personal consumer information is its bread and butter and probably will be for the foreseeable future. In a world where all the information ever created by humans is quickly becoming instantly available, advertising continues to be the dominant profit paradigm.
While using Google metrics to target consumers in offline markets – which is exactly what happened -- may sound like a good deal for offline advertisers, those methods of communicating with consumers are dying. The metrics Google uses to perfectly identify the right market for ad placements online just didn't translate to the offline world. Tracking the success of Google's ad placements proved difficult, and both radio and print executives were reluctant to turn over their advertising systems to Google's methods [source: Vascellaro].
Of course, Google's private and personal consumer information is its bread and butter and probably will be for the foreseeable future. In a world where all the information ever created by humans is quickly becoming instantly available, advertising continues to be the dominant profit paradigm.
While using Google metrics to target consumers in offline markets – which is exactly what happened -- may sound like a good deal for offline advertisers, those methods of communicating with consumers are dying. The metrics Google uses to perfectly identify the right market for ad placements online just didn't translate to the offline world. Tracking the success of Google's ad placements proved difficult, and both radio and print executives were reluctant to turn over their advertising systems to Google's methods [source: Vascellaro].
Screen capture by HowStuffWorks staff
In 2005, two Google acquisitions in particular stick out: Android,
and Dodgeball. Android, of course, has no place being discussed in this
article. But Dodgeball is more interesting, as the first case on our
list of Google acquiring and developing an idea that eventually
succeeded elsewhere, becoming the current standard.
Dodgeball was a location-specific social networking site and was acquired, along with cofounder Dennis Crowley, in May 2005 [source: Seigler]. Again, we see the forward-thinking merge between online and real-world life, as applications like this use smartphone technology to connect us, tout our social experiences and favorite locations, and send out all manner of food portraiture to everyone we know. Perfect Google situation, right?
So what happened? Well, nothing. For two years, that is, until Crowley left Google in frustration and founded Foursquare. The blame here rests in the fact that the idea was too prescient, that the hardware took too long to catch up to the idea, but catch up it did. Now, of course, Google's got Latitude, and Facebook's Places may take the Foursquare crown as the check-in app of choice.
Of course, neither of those latter apps have what made Foursquare such a hit -- the gamification aspect, in which demonstrated loyalty to a given business or location results in various badges and bells -- but if we follow our "real world parallel" model, it seems those extra features won't really matter as much moving forward.
Users check in now because that's just what you do. It's not to get a virtual treat; checking in is faster and easier than tweeting or Facebooking our location to our friends. And with location mapping becoming a standard part of photo apps like Instagram, the concept of the check-in itself has morphed itself into closer approximation of what the connected life has become: The augmentation, rather than the replacement, of reality.
Jaiku
Dodgeball was a location-specific social networking site and was acquired, along with cofounder Dennis Crowley, in May 2005 [source: Seigler]. Again, we see the forward-thinking merge between online and real-world life, as applications like this use smartphone technology to connect us, tout our social experiences and favorite locations, and send out all manner of food portraiture to everyone we know. Perfect Google situation, right?
So what happened? Well, nothing. For two years, that is, until Crowley left Google in frustration and founded Foursquare. The blame here rests in the fact that the idea was too prescient, that the hardware took too long to catch up to the idea, but catch up it did. Now, of course, Google's got Latitude, and Facebook's Places may take the Foursquare crown as the check-in app of choice.
Of course, neither of those latter apps have what made Foursquare such a hit -- the gamification aspect, in which demonstrated loyalty to a given business or location results in various badges and bells -- but if we follow our "real world parallel" model, it seems those extra features won't really matter as much moving forward.
Users check in now because that's just what you do. It's not to get a virtual treat; checking in is faster and easier than tweeting or Facebooking our location to our friends. And with location mapping becoming a standard part of photo apps like Instagram, the concept of the check-in itself has morphed itself into closer approximation of what the connected life has become: The augmentation, rather than the replacement, of reality.
Jaiku
In 2005, two Google acquisitions in particular stick out: Android,
and Dodgeball. Android, of course, has no place being discussed in this
article. But Dodgeball is more interesting, as the first case on our
list of Google acquiring and developing an idea that eventually
succeeded elsewhere, becoming the current standard.
Dodgeball was a location-specific social networking site and was acquired, along with cofounder Dennis Crowley, in May 2005 [source: Seigler]. Again, we see the forward-thinking merge between online and real-world life, as applications like this use smartphone technology to connect us, tout our social experiences and favorite locations, and send out all manner of food portraiture to everyone we know. Perfect Google situation, right?
So what happened? Well, nothing. For two years, that is, until Crowley left Google in frustration and founded Foursquare. The blame here rests in the fact that the idea was too prescient, that the hardware took too long to catch up to the idea, but catch up it did. Now, of course, Google's got Latitude, and Facebook's Places may take the Foursquare crown as the check-in app of choice.
Of course, neither of those latter apps have what made Foursquare such a hit -- the gamification aspect, in which demonstrated loyalty to a given business or location results in various badges and bells -- but if we follow our "real world parallel" model, it seems those extra features won't really matter as much moving forward.
Users check in now because that's just what you do. It's not to get a virtual treat; checking in is faster and easier than tweeting or Facebooking our location to our friends. And with location mapping becoming a standard part of photo apps like Instagram, the concept of the check-in itself has morphed itself into closer approximation of what the connected life has become: The augmentation, rather than the replacement, of reality.
Dodgeball was a location-specific social networking site and was acquired, along with cofounder Dennis Crowley, in May 2005 [source: Seigler]. Again, we see the forward-thinking merge between online and real-world life, as applications like this use smartphone technology to connect us, tout our social experiences and favorite locations, and send out all manner of food portraiture to everyone we know. Perfect Google situation, right?
So what happened? Well, nothing. For two years, that is, until Crowley left Google in frustration and founded Foursquare. The blame here rests in the fact that the idea was too prescient, that the hardware took too long to catch up to the idea, but catch up it did. Now, of course, Google's got Latitude, and Facebook's Places may take the Foursquare crown as the check-in app of choice.
Of course, neither of those latter apps have what made Foursquare such a hit -- the gamification aspect, in which demonstrated loyalty to a given business or location results in various badges and bells -- but if we follow our "real world parallel" model, it seems those extra features won't really matter as much moving forward.
Users check in now because that's just what you do. It's not to get a virtual treat; checking in is faster and easier than tweeting or Facebooking our location to our friends. And with location mapping becoming a standard part of photo apps like Instagram, the concept of the check-in itself has morphed itself into closer approximation of what the connected life has become: The augmentation, rather than the replacement, of reality.
Screen capture by HowStuffWorks staff
Google acquired microblogging site Jaiku in October 2007, but by
January 2009, it was clear that Twitter was the official winner in the
short post race [source: Kincaid].
A social network is only as powerful as the users themselves, and
Twitter was already well on its way by the time of this acquisition.
The divorce between Google and Jaiku is surrounded by rumors of internal bad blood, but either way, this Finnish import -- so-named because the microblogging aspect makes its messages look like haiku -- was open-source after 2009. In 2011, Google announced that it was shutting down Jaiku for good, effective Jan. 15, 2012 [source: Horowitz]. Perhaps in the same way that the MySpace graveyard has over the years become a home-base headquarters for smaller unknown bands -- a development presaged by MySpace's music-integration technology, which still sets it apart from most social networks that aren't actively concerned with music -- it could have become something new. Now, we'll never know.
Google Notebook and Shared Stuff
The divorce between Google and Jaiku is surrounded by rumors of internal bad blood, but either way, this Finnish import -- so-named because the microblogging aspect makes its messages look like haiku -- was open-source after 2009. In 2011, Google announced that it was shutting down Jaiku for good, effective Jan. 15, 2012 [source: Horowitz]. Perhaps in the same way that the MySpace graveyard has over the years become a home-base headquarters for smaller unknown bands -- a development presaged by MySpace's music-integration technology, which still sets it apart from most social networks that aren't actively concerned with music -- it could have become something new. Now, we'll never know.
Google Notebook and Shared Stuff
Google acquired microblogging site Jaiku in October 2007, but by
January 2009, it was clear that Twitter was the official winner in the
short post race [source: Kincaid].
A social network is only as powerful as the users themselves, and
Twitter was already well on its way by the time of this acquisition.
The divorce between Google and Jaiku is surrounded by rumors of internal bad blood, but either way, this Finnish import -- so-named because the microblogging aspect makes its messages look like haiku -- was open-source after 2009. In 2011, Google announced that it was shutting down Jaiku for good, effective Jan. 15, 2012 [source: Horowitz]. Perhaps in the same way that the MySpace graveyard has over the years become a home-base headquarters for smaller unknown bands -- a development presaged by MySpace's music-integration technology, which still sets it apart from most social networks that aren't actively concerned with music -- it could have become something new. Now, we'll never know.
The divorce between Google and Jaiku is surrounded by rumors of internal bad blood, but either way, this Finnish import -- so-named because the microblogging aspect makes its messages look like haiku -- was open-source after 2009. In 2011, Google announced that it was shutting down Jaiku for good, effective Jan. 15, 2012 [source: Horowitz]. Perhaps in the same way that the MySpace graveyard has over the years become a home-base headquarters for smaller unknown bands -- a development presaged by MySpace's music-integration technology, which still sets it apart from most social networks that aren't actively concerned with music -- it could have become something new. Now, we'll never know.
Screen capture by HowStuffWorks staff
While Google Docs
has become the shared-document service that Google Wave (more on this
coming up) partially wanted to be, the company's never come up with an
application that can compete with apps like Evernote for the
getting-things-done crowd. Cutting and pasting clips that retain their
Web citation seems like a sure thing -- especially when integrated with
the browser itself -- which is why Google has tried it so many times.
And even so, at the end of the day, the learning curve or feature load has either been too high, or the interface has been too clunky. The world of squirreled-away factlets and quotations remains firmly under the regime of those app developers with the leanest extensions and the simplest features. When everything's in the cloud, being able to port your notes and tasks and links from home to phone to office is no longer a selling point. (And again, we see the theme of seamless integration of the tech until you can't see it anymore.)
Likewise, the improbably named Shared Stuff tried to work the Google Docs and Google Notebook angles by making those clips and notes available to everybody [source: Pash]. The development had problems -- it's been called buggy, and it never really integrated into the Google world -- but the result was just a less-fun version of social bookmarking sites like Delicio.us, which privileged the "social" aspect of the concept into its own activity: Social bookmarking is exactly what it sounds like, whether it takes the form of Delicio.us, Reddit or even BuzzFeed. What's important isn't so much what you share, but what you and your friends have to say about it. (Those are the aspects of Notebook and Shared Stuff that were integrated into Google Reader.)
Google Buzz
And even so, at the end of the day, the learning curve or feature load has either been too high, or the interface has been too clunky. The world of squirreled-away factlets and quotations remains firmly under the regime of those app developers with the leanest extensions and the simplest features. When everything's in the cloud, being able to port your notes and tasks and links from home to phone to office is no longer a selling point. (And again, we see the theme of seamless integration of the tech until you can't see it anymore.)
Likewise, the improbably named Shared Stuff tried to work the Google Docs and Google Notebook angles by making those clips and notes available to everybody [source: Pash]. The development had problems -- it's been called buggy, and it never really integrated into the Google world -- but the result was just a less-fun version of social bookmarking sites like Delicio.us, which privileged the "social" aspect of the concept into its own activity: Social bookmarking is exactly what it sounds like, whether it takes the form of Delicio.us, Reddit or even BuzzFeed. What's important isn't so much what you share, but what you and your friends have to say about it. (Those are the aspects of Notebook and Shared Stuff that were integrated into Google Reader.)
Google Buzz
While Google Docs
has become the shared-document service that Google Wave (more on this
coming up) partially wanted to be, the company's never come up with an
application that can compete with apps like Evernote for the
getting-things-done crowd. Cutting and pasting clips that retain their
Web citation seems like a sure thing -- especially when integrated with
the browser itself -- which is why Google has tried it so many times.
And even so, at the end of the day, the learning curve or feature load has either been too high, or the interface has been too clunky. The world of squirreled-away factlets and quotations remains firmly under the regime of those app developers with the leanest extensions and the simplest features. When everything's in the cloud, being able to port your notes and tasks and links from home to phone to office is no longer a selling point. (And again, we see the theme of seamless integration of the tech until you can't see it anymore.)
Likewise, the improbably named Shared Stuff tried to work the Google Docs and Google Notebook angles by making those clips and notes available to everybody [source: Pash]. The development had problems -- it's been called buggy, and it never really integrated into the Google world -- but the result was just a less-fun version of social bookmarking sites like Delicio.us, which privileged the "social" aspect of the concept into its own activity: Social bookmarking is exactly what it sounds like, whether it takes the form of Delicio.us, Reddit or even BuzzFeed. What's important isn't so much what you share, but what you and your friends have to say about it. (Those are the aspects of Notebook and Shared Stuff that were integrated into Google Reader.)
And even so, at the end of the day, the learning curve or feature load has either been too high, or the interface has been too clunky. The world of squirreled-away factlets and quotations remains firmly under the regime of those app developers with the leanest extensions and the simplest features. When everything's in the cloud, being able to port your notes and tasks and links from home to phone to office is no longer a selling point. (And again, we see the theme of seamless integration of the tech until you can't see it anymore.)
Likewise, the improbably named Shared Stuff tried to work the Google Docs and Google Notebook angles by making those clips and notes available to everybody [source: Pash]. The development had problems -- it's been called buggy, and it never really integrated into the Google world -- but the result was just a less-fun version of social bookmarking sites like Delicio.us, which privileged the "social" aspect of the concept into its own activity: Social bookmarking is exactly what it sounds like, whether it takes the form of Delicio.us, Reddit or even BuzzFeed. What's important isn't so much what you share, but what you and your friends have to say about it. (Those are the aspects of Notebook and Shared Stuff that were integrated into Google Reader.)
The first thing Google Buzz did wrong was sneak up on users. In February of 2010, it was automatically added to Gmail, as an opt-outservice that sneakily appeared as a folder in the comfy old Inbox without warning.
So what was inside that spooky new folder? It was Google Reader, in essence, which was a great experience during the time it was most-used -- before, that is, RSS as a Web standard gave way to personally-curated tablet readers (including Google Currents) and similar app-based ways of keeping track of our favorite sites. Which was a transition that was already underway when Buzz appeared, so Google's initiative basically amounted to (or would, over the next year) just another folder with a continually rising "Unread" count, with all the subconscious stress that entails.
Perhaps if Google Buzz had incorporated some sort of reward for getting through those -- formerly enjoyable! -- updates from our favorite sites, it would have done better. In any case, the tablet revolution has brought the cycle back around: Now, we read magazines on something shaped like a magazine, rather than reading blog posts on something shaped like our e-mail. In late 2011, Google put Buzz out to pasture [source: Wasserman].
Wikipedia Alternatives
So what was inside that spooky new folder? It was Google Reader, in essence, which was a great experience during the time it was most-used -- before, that is, RSS as a Web standard gave way to personally-curated tablet readers (including Google Currents) and similar app-based ways of keeping track of our favorite sites. Which was a transition that was already underway when Buzz appeared, so Google's initiative basically amounted to (or would, over the next year) just another folder with a continually rising "Unread" count, with all the subconscious stress that entails.
Perhaps if Google Buzz had incorporated some sort of reward for getting through those -- formerly enjoyable! -- updates from our favorite sites, it would have done better. In any case, the tablet revolution has brought the cycle back around: Now, we read magazines on something shaped like a magazine, rather than reading blog posts on something shaped like our e-mail. In late 2011, Google put Buzz out to pasture [source: Wasserman].
The past can get a little fuzzy, but most of us remember that time
right before Wikipedia's debut when "wiki" was its own dominant concept.
Television
fandoms and other information-rich communities still maintain wikis full
of user-edited and -confirmed facts about the things that they love.
What makes Wikipedia special is the size and devotion of its community;
despite what your high school English teacher has to say about it, the
fact that "anybody" can edit Wikipedia pages doesn't necessarily make
the information invalid. All accepted knowledge is written by committee,
as they say.
And what does that have to do with Google? SearchWiki, Knol and SideWiki, that's what. A whole sequence of Wikipedia add-ons and alternatives, developed by Google since the summer of 2008. Couldn't beat 'em (Knol, a collection of user-written articles), couldn't join 'em (SearchWiki, which enabled users to sort and annotate search results), finally gave up (SideWiki, a browser extension to annotate Web pages).
Any attempt at a "Wikipedia killer" -- even one administrated by user-beloved Google -- was never going to measure up in sheer crowdsourcing power, and Knol shuttered in May of 2012 [source: Albanesius]. Perhaps if there had been any outstanding problems in the Wikipedia interface, Knol would have had a shot, but the fact is that Wikipedia's pretty solid, offering enough usefulness to every level of user -- from the novice, to experts on the very subjects they're reading about, which is pretty amazing if you think about it -- that everyone is welcome both to search and to provide the information being searched, often at once.
As for SearchWiki, users seemed reluctant to mess with Google's organic search results, so it was replaced with a star system in late 2010. [source: Dupont]. In SideWiki's case, users never really took to the use of a sidebar to comment on Web pages, and Google pulled the plug in September 2011 [source: Eustace]
So what was inside that spooky new folder? It was Google Reader, in essence, which was a great experience during the time it was most-used -- before, that is, RSS as a Web standard gave way to personally-curated tablet readers (including Google Currents) and similar app-based ways of keeping track of our favorite sites. Which was a transition that was already underway when Buzz appeared, so Google's initiative basically amounted to (or would, over the next year) just another folder with a continually rising "Unread" count, with all the subconscious stress that entails.
Perhaps if Google Buzz had incorporated some sort of reward for getting through those -- formerly enjoyable! -- updates from our favorite sites, it would have done better. In any case, the tablet revolution has brought the cycle back around: Now, we read magazines on something shaped like a magazine, rather than reading blog posts on something shaped like our e-mail. In late 2011, Google put Buzz out to pasture [source: Wasserman].
Wikipedia Alternatives
So what was inside that spooky new folder? It was Google Reader, in essence, which was a great experience during the time it was most-used -- before, that is, RSS as a Web standard gave way to personally-curated tablet readers (including Google Currents) and similar app-based ways of keeping track of our favorite sites. Which was a transition that was already underway when Buzz appeared, so Google's initiative basically amounted to (or would, over the next year) just another folder with a continually rising "Unread" count, with all the subconscious stress that entails.
Perhaps if Google Buzz had incorporated some sort of reward for getting through those -- formerly enjoyable! -- updates from our favorite sites, it would have done better. In any case, the tablet revolution has brought the cycle back around: Now, we read magazines on something shaped like a magazine, rather than reading blog posts on something shaped like our e-mail. In late 2011, Google put Buzz out to pasture [source: Wasserman].
Screen capture by HowStuffWorks staff
And what does that have to do with Google? SearchWiki, Knol and SideWiki, that's what. A whole sequence of Wikipedia add-ons and alternatives, developed by Google since the summer of 2008. Couldn't beat 'em (Knol, a collection of user-written articles), couldn't join 'em (SearchWiki, which enabled users to sort and annotate search results), finally gave up (SideWiki, a browser extension to annotate Web pages).
Any attempt at a "Wikipedia killer" -- even one administrated by user-beloved Google -- was never going to measure up in sheer crowdsourcing power, and Knol shuttered in May of 2012 [source: Albanesius]. Perhaps if there had been any outstanding problems in the Wikipedia interface, Knol would have had a shot, but the fact is that Wikipedia's pretty solid, offering enough usefulness to every level of user -- from the novice, to experts on the very subjects they're reading about, which is pretty amazing if you think about it -- that everyone is welcome both to search and to provide the information being searched, often at once.
As for SearchWiki, users seemed reluctant to mess with Google's organic search results, so it was replaced with a star system in late 2010. [source: Dupont]. In SideWiki's case, users never really took to the use of a sidebar to comment on Web pages, and Google pulled the plug in September 2011 [source: Eustace]
Google Video
The past can get a little fuzzy, but most of us remember that time
right before Wikipedia's debut when "wiki" was its own dominant concept.
Television
fandoms and other information-rich communities still maintain wikis full
of user-edited and -confirmed facts about the things that they love.
What makes Wikipedia special is the size and devotion of its community;
despite what your high school English teacher has to say about it, the
fact that "anybody" can edit Wikipedia pages doesn't necessarily make
the information invalid. All accepted knowledge is written by committee,
as they say.
And what does that have to do with Google? SearchWiki, Knol and SideWiki, that's what. A whole sequence of Wikipedia add-ons and alternatives, developed by Google since the summer of 2008. Couldn't beat 'em (Knol, a collection of user-written articles), couldn't join 'em (SearchWiki, which enabled users to sort and annotate search results), finally gave up (SideWiki, a browser extension to annotate Web pages).
Any attempt at a "Wikipedia killer" -- even one administrated by user-beloved Google -- was never going to measure up in sheer crowdsourcing power, and Knol shuttered in May of 2012 [source: Albanesius]. Perhaps if there had been any outstanding problems in the Wikipedia interface, Knol would have had a shot, but the fact is that Wikipedia's pretty solid, offering enough usefulness to every level of user -- from the novice, to experts on the very subjects they're reading about, which is pretty amazing if you think about it -- that everyone is welcome both to search and to provide the information being searched, often at once.
As for SearchWiki, users seemed reluctant to mess with Google's organic search results, so it was replaced with a star system in late 2010. [source: Dupont]. In SideWiki's case, users never really took to the use of a sidebar to comment on Web pages, and Google pulled the plug in September 2011 [source: Eustace].
And what does that have to do with Google? SearchWiki, Knol and SideWiki, that's what. A whole sequence of Wikipedia add-ons and alternatives, developed by Google since the summer of 2008. Couldn't beat 'em (Knol, a collection of user-written articles), couldn't join 'em (SearchWiki, which enabled users to sort and annotate search results), finally gave up (SideWiki, a browser extension to annotate Web pages).
Any attempt at a "Wikipedia killer" -- even one administrated by user-beloved Google -- was never going to measure up in sheer crowdsourcing power, and Knol shuttered in May of 2012 [source: Albanesius]. Perhaps if there had been any outstanding problems in the Wikipedia interface, Knol would have had a shot, but the fact is that Wikipedia's pretty solid, offering enough usefulness to every level of user -- from the novice, to experts on the very subjects they're reading about, which is pretty amazing if you think about it -- that everyone is welcome both to search and to provide the information being searched, often at once.
As for SearchWiki, users seemed reluctant to mess with Google's organic search results, so it was replaced with a star system in late 2010. [source: Dupont]. In SideWiki's case, users never really took to the use of a sidebar to comment on Web pages, and Google pulled the plug in September 2011 [source: Eustace].
Google Video attempted to crush YouTube
using merely its beautifully lean interface, its whipsmart programming
... and the complete lack of any need for something that already exists.
Again, we see the crowdsourced chaos of a Wikipedia in YouTube, with
user-administrated levels of appreciation and reputation bringing the
cream to the surface. While Google Videos (plural, totally different
name), Google Video's successor, is still a storehouse for certain video
streams, it's taken the more tightly curated route of sites like its
early partner Vimeo. And of course, Google eventually bought YouTube
anyway, to the tune of $1.65 billion in stock. So it all worked out.
The story of Google Video isn't merely that of an unprovoked attack on an Internet behemoth, though. The truth is much stranger. In January 2005, the roots of what would become Google Video first debuted, turning television broadcasts into searchable transcripts. By summer of that year, they started supporting video uploads and sharing, and by the end of its first year of life it has lost the original transcript idea altogether (although as of 2012, it's available for some video on YouTube, which implies Google's not done entirely with this concept).
Whatever slim chance the site might have had, whatever improvements or fun user-experience innovations that might have put it over the top (like Facebook's clean interface did, for example, once upon a time), Google Video decided to go another way: by introducing a proprietary file type and player, drastically increasing the amount of "stuff" you had to do in order to create or enjoy content on the site. Sometimes this works -- all file extensions and media players came from somewhere, right? -- but it's not a great strategy when you've started a fight with a perfectly useable site like YouTube, whose popularity has already made it the standard. And certainly not when portability between devices and screens had already become the new measure of a killer app.
After the YouTube acquisition, and having failed at becoming the rebranded name of the service, Google Video changed shape once again, this time into a video rental service (once again, heading into competition with the guy that already won, in this case Netflix). Now it's back to its form as a YouTube analogue -- which is good news to anybody who already has content hosted there. A static collection of videos, now that they've disabled uploading, will stand as testament to the brief time Google Video filled a need -- over a billion dollars later -- for its community. At least until they've folded back into YouTube, presumably.
The story of Google Video isn't merely that of an unprovoked attack on an Internet behemoth, though. The truth is much stranger. In January 2005, the roots of what would become Google Video first debuted, turning television broadcasts into searchable transcripts. By summer of that year, they started supporting video uploads and sharing, and by the end of its first year of life it has lost the original transcript idea altogether (although as of 2012, it's available for some video on YouTube, which implies Google's not done entirely with this concept).
Whatever slim chance the site might have had, whatever improvements or fun user-experience innovations that might have put it over the top (like Facebook's clean interface did, for example, once upon a time), Google Video decided to go another way: by introducing a proprietary file type and player, drastically increasing the amount of "stuff" you had to do in order to create or enjoy content on the site. Sometimes this works -- all file extensions and media players came from somewhere, right? -- but it's not a great strategy when you've started a fight with a perfectly useable site like YouTube, whose popularity has already made it the standard. And certainly not when portability between devices and screens had already become the new measure of a killer app.
After the YouTube acquisition, and having failed at becoming the rebranded name of the service, Google Video changed shape once again, this time into a video rental service (once again, heading into competition with the guy that already won, in this case Netflix). Now it's back to its form as a YouTube analogue -- which is good news to anybody who already has content hosted there. A static collection of videos, now that they've disabled uploading, will stand as testament to the brief time Google Video filled a need -- over a billion dollars later -- for its community. At least until they've folded back into YouTube, presumably.
Google Wave
Perhaps the most famous Google failure, Wave also bears the
distinction of being the biggest Google failure. A collection of
unnecessary features bundled together in unnecessary -- and often
bewildering – ways, Google Wave tried to be everything to everyone in
terms of content sharing, in the same way that Google+
is attempting to take over the social realm. And while it's not yet
certain whether Google+ will flatline, the time to mourn Wave has come
and gone.
Want to send an e-mail? You already have Gmail, but if for some reason you'd like to send that e-mail to a hard-to-understand list of people through a counterintuitive process, Wave can help. Would you like to turn that e-mail into a song, or a video, or a conversation about songs and videos that itself contains and is made of those things? Want to juggle people coming in and out of that conversation, never quite sure to whom you're talking or whether they've been following the conversation the whole time? Want the always-on capability to form sidebar conversations alongside the main conversation, creating a constant -- and possibly valid -- paranoia that everybody is talking about you behind your back? Would you like to take all the most irritating lags and social awkwardness of chat rooms and combine them -- along with the worst things about online document-collaboration, online flame wars and awkward parties where your work friends meet your regular friends -- into a single application that none of those people know how to use either?
Of course, it wasn't really that bad. What gets left out of this story is the fact that -- like most Apple products, like most presidents -- the anticipation of a product release can easily overshadow any actual value. If we've paid enough money and gotten enough usefulness out of a so-so project, we'll swear the Emperor is wearing the best clothes in town. But if the product is free, or we feel defeated by it, then it becomes the worst thing that has ever happened.
Google Wave is no different. It made its debut through the "invite" system that was in vogue in 2009, like the tremendous Google Voice, and like Voice, it spread into the culture through the people most likely to turn backflips on release day, and of course most likely to talk about it for at least the two weeks either side. A risky strategy, for a project with no broad-spectrum use that would take more than those two weeks to learn, even for a hard-headed Google fanatic. Even as performance art, or a joke.
The fact is, even seasoned programmers can have trouble explaining to the layman why Wave was so unloved. Part of it is the complexity of code language, the precise reasons that it failed to integrate with other Google features and suites, that don't enter into here. And part -- likely most -- of it is that anticipation-backlash effect. But perhaps the "right place, wrong time" aspect is also in play. Whatever features users liked in Wave will likely make their way into a future project or acquisition. Those pieces of the broken and abandoned products that make up Google's Island of Misfit Toys can always be picked up, dusted off and integrated into a new configuration.
source:http://computer.howstuffworks.com
Want to send an e-mail? You already have Gmail, but if for some reason you'd like to send that e-mail to a hard-to-understand list of people through a counterintuitive process, Wave can help. Would you like to turn that e-mail into a song, or a video, or a conversation about songs and videos that itself contains and is made of those things? Want to juggle people coming in and out of that conversation, never quite sure to whom you're talking or whether they've been following the conversation the whole time? Want the always-on capability to form sidebar conversations alongside the main conversation, creating a constant -- and possibly valid -- paranoia that everybody is talking about you behind your back? Would you like to take all the most irritating lags and social awkwardness of chat rooms and combine them -- along with the worst things about online document-collaboration, online flame wars and awkward parties where your work friends meet your regular friends -- into a single application that none of those people know how to use either?
Of course, it wasn't really that bad. What gets left out of this story is the fact that -- like most Apple products, like most presidents -- the anticipation of a product release can easily overshadow any actual value. If we've paid enough money and gotten enough usefulness out of a so-so project, we'll swear the Emperor is wearing the best clothes in town. But if the product is free, or we feel defeated by it, then it becomes the worst thing that has ever happened.
Google Wave is no different. It made its debut through the "invite" system that was in vogue in 2009, like the tremendous Google Voice, and like Voice, it spread into the culture through the people most likely to turn backflips on release day, and of course most likely to talk about it for at least the two weeks either side. A risky strategy, for a project with no broad-spectrum use that would take more than those two weeks to learn, even for a hard-headed Google fanatic. Even as performance art, or a joke.
The fact is, even seasoned programmers can have trouble explaining to the layman why Wave was so unloved. Part of it is the complexity of code language, the precise reasons that it failed to integrate with other Google features and suites, that don't enter into here. And part -- likely most -- of it is that anticipation-backlash effect. But perhaps the "right place, wrong time" aspect is also in play. Whatever features users liked in Wave will likely make their way into a future project or acquisition. Those pieces of the broken and abandoned products that make up Google's Island of Misfit Toys can always be picked up, dusted off and integrated into a new configuration.
source:http://computer.howstuffworks.com
No comments:
Post a Comment