Google Real-Time Search

» see google.com/realtime as it lived, on the wayback machine

~ the obituary ~

Google built this thing on top of Twitter, and in 2010 Twitter had barely arrived in Turkey. The country was still on Facebook and had not fully let go of MSN. The Turkish rush onto Twitter came a couple of years later, well after this feature was already dead. So it sat in my search results the way it sat in everyone's, one more word in the left sidebar, and I had no particular reason to press it.

It was announced on 7 December 2009 and it did something a search engine had never done. Below the ordinary blue links, Google opened a window onto posts that were seconds old and let them keep arriving while you watched. Twitter first, with blogs and news sites alongside it, and Facebook status updates added from 24 February 2010 onward. You could pause the flow and start it again. The rest of the page stayed as static as a page has always been, and this one panel would not stop moving.

The whole idea was a response to Bing. Microsoft had signed its own deal with Twitter and was already showing posts in results, and Google, which had spent a decade insisting that the right answer took a few hours of crawling to find, suddenly needed an answer that was thirty seconds old. It got one by writing a check.

~ the case file ~

~ the river ~

The interface was better than the feature deserved. Posts descended down the middle of the page in what Google's own people called a river. On the right there was a bar chart showing how often your search term was being posted, and under that a list of the most reposted items and the links people were sharing. Hashtag queries worked, and a retweeted image arrived as a thumbnail rather than a URL. Promoted tweets sat pinned above the river, which is the one part of the design that has aged into something familiar.

In January 2011 it learned where posts came from, adding geolocation next to results, and it learned to look inside shortened links so you could query the destination rather than the wrapper. That was the high point. Six months later the whole thing was a 404.

~ the deal ~

Google had signed with Twitter in October 2009 for access to a special feed of updates, and the agreement was time-limited in a way almost nobody outside the two companies knew. On 2 July 2011 it expired. The Real-Time link disappeared from the sidebar and google.com/realtime started returning an error.

Google's first public explanation blamed the launch of Google+, which had gone live four days earlier, and suggested the outage was temporary. The real reason came out afterward and was much duller. "Since October of 2009, we have had an agreement with Twitter to include their updates in our search results through a special feed, and that agreement expired on July 2." A feature that felt like the future of search turned out to be a rental.

~ that's all we know ~

The company said the service would return. "Our vision is to have google.com/realtime include Google+ information along with other realtime data from a variety of sources." No date was given. None was ever given. Google+ lasted another eight years and Real-Time Search was never rebuilt into it, and by the time Twitter would have sold the feed again the two companies had a different relationship entirely.

The Wayback Machine has the last working capture on 25 June 2011 and a 404 on 29 August. The archived error page is Google's standard one, written in the deadpan voice they used for broken URLs in those years, and it reads: "404. That's an error. The requested URL /Realtime was not found on this server. That's all we know." Today the address answers with a redirect to plain search, which is the only honest thing left to do with it.

@}-->--

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

« previous  ·  random  ·  next »