Larry Page: Twitter made Google focus on Realtime Search
  • 154 Comments
by Mike Butcher on May 19, 2009

Larry Page and Eric Schmidt from Google did a double header interview from the Google Zeitgeist conference just outside of London today. But the real question on our lips was what is Google going to do about the astounding buzz around realtime search and Twitter?

During a press conference I asked the question of of of Google’s executives, and the answer came back that “the kind of innovation like what Twitter is doing and what we’re doing is increasing search speed, relevance , freshness and comprehensiveness. Other companies will come up with solutions of course.”

Not a great answer.

Luckily, Loic Le Meur is also here and put Larry on the spot on stage, and captured:

“I have always thought we needed to index the web every second to allow real time search. At first, my team laughed and did not believe me. With Twitter, now they know they have to do it. Not everybody needs sub-second indexing but people are getting pretty excited about realtime.”

See, Larry actually came up with the idea first.

Responses

Comments rss icon

  • Is there a video for this? If so can you please post it?

  • I don’t think I “got” the power of real-time search until a few nights ago when I saw a few “earthquake” tweets. After a few searches I had an idea of the location, size and damage of the California earthquake literally seconds after it happened.

    I then checked every news site and did a Google search…nothing (obviously)

    • The best is a combination of Google and Twitter search like http://www.Yauba.com or http://www.Oneriot.com

      I am sure both Google and Twitter will evolve in that direction.

      • For the last time, there is nothing new about a real time stream.

        Reuters, Bloomberg, CNN have been doing real time news streams for years.

        I have been able to search from my Bloomberg terminal from time immemorial. All the new real time search engines are really doing things that these companies did years ago.

        • Sainsbury, real-time search/streams is very new because it is coming in from everybody, rather than after the fact by journalists. Sure news stories are always coming in, but this is like getting the news live as the event happens. Prior to this, there simply wasn’t enough sharing occurring for it to resemble real-time.

        • “All the new real time search engines are really doing things that these companies did years ago.”

          I disagree.

          Twitter, Facebook, et al provide a massive, raw stream of news, events, ideas, photos, videos that can be tapped into by anyone at any time. From large scale news to hyper-local, it’s all in the stream and can be searched/filtered as needed. Very powerful stuff that is just beginning to be harnessed.

          In my opinion that is very different from a “live” stream of delayed (vetted…?), filtered and spun CNN/Reuters news, if only because it removes a major bottle-neck from the process.

      • There’s also this one that is simple and seems to work well-

        http://www.searchmotive.com

    • Yeah, earthquakes, terrorist attacks etc. But so what? Who cares….there’s no business plan in this, it’s just a gimmick where you hear about something on Twitter a minute before CNN on the TV or whatever – Wooooowwww. Or you can see if Gmail is down for you or everyone. Again, wow.

      Really is that it? Is that the point of realtime?

      Can anyone come up with other ideas that will make money?

    • Right, an up to the second notification on every meaningless occurance. You should stop wasting your time and start contributing to the real world.

  • theres a reason the founders of googl are hidden from the public and not in the spotlight often. the more they speak the less sense they make.

    LatitudeLocator.com – find yourself

  • Interesting to see what happens, when it comes to real time search Google is wayyy behind Twitter right now, and everyone agree “real time” is the future of the Internet

    • Well, to say they are wayyy behind is someone of an exaggeration.

      Twitter does not offer real time search.

      Twitter offers twitter search.

      Indeed, none of the so called real time search engines from itpints to scoopler to yauba offer real real time search.

      They offer twitter search.

      Real, real time search would be able to index, sort, and rank through everythng posted right now and determine its relevance to a given query.

      This is very different than querying your own database and returning results and calling that real time.

      If anyone can do it, it is a company with strong search technologies like Google or aggregation capabilities like Friendfeed or auto classification technologies like Yauba.

      Twitter search is searching twitter, that’s it.

      • Salladin said…
        Real, real time search would be able to index, sort, and rank through everythng posted right now and determine its relevance to a given query.

        …Twitter search is searching twitter, that’s it.

        Well said Salladin.

  • Great article. I would like better answers from the founders.

  • A real-time search is obviously a good idea. But that is feasible only through Twitter now. Does Larry’s answer imply anything about Google’s willingness to acquire Twitter?

    • Windowslogy said…
      A real-time search is obviously a good idea.

      You can’t do real-time PageRank computation, because it is technologically impossible to do it today. You can do it if one has built a quantum computer to do the number-crunching. Google or anyone else can do what Twitter is doing, because you don’t need massive number-crunching to index anything. Google is not going to do that (ie, replicate what Twitter is doing) or otherwise they’ll have to ditch PageRank, but the chance of them dropping PageRank is zero. I guess that they will try to extend PageRank to include a time-stamp dimension in one computational framework to bring in document recency where PageRank fails, but then again, it is going to be undoable to compute it in real-time.

  • Dear Larry Page i love your work and want to know how you going to but twiiter or is it going to stay the employee of the month for long time working for google and is indexing

  • Twitter search is about knowing what people talk about *right now* – Google search is about what people talk / think about yesterday … last week … last year.

    I bet Google is breaking a sweat over the fact that they don’t ‘own’ the real time web – twitter does. I’d love to know what happened when they made an offer on twitter. Now they don’t own the most important data source for real time and making their search work with it will be so much harder and more expensive. Good times :)

  • twitter is not tackling real time search. twitter is tackling real time twitter search – very, very different.

    larry is talking about something much larger, much more complex. duh.

    take off your twitter goggles for a second, and you may be able to see that you (and the rest of the team) are playing the part of pawn. its sad to watch.

  • Yeah, i agree, we need a real time environment…

  • This was all over twitter last week. Remind me again, why do we need TC?

  • If Google can index in real time it would make for some interesting side effects with websites. Basically your website would be a Twitter platform. Currently, unless you are a gigantic and very active website such as Tech Crunch, it can take an awful long time for Google to index new content. It can take even longer for Google to show this new content in searches. So trying to break news with your website is often difficult because the news can be there but no one can see it in searches. I think this is what they are getting at with this announcement.

    • it would require a push. and all google would need to do is ask for it…it would become seo best practice within 6 months.

      • I think it would take an awful lot more computer power to do though. Consistency would be a must and that I think might require overkill to make sure. Either way it might be some logistical hurdles to get over. I am pretty sure they reason they dont do it now is because of the resources it takes.

        Another issue to consider is that a lot of sites cant handle the traffic that Googlebots exert on a site. It might take web hosts to enable more bandwith to even handle the load.

  • Very interesting!

    And it looks like they’ve got their work cut out for them. We’ve got over 5mil items in our system, and even with sitemap XMLs, their crawlers have not indexed them all yet.

    I’m not sure that I’d want them to index faster – I don’t want to pay for 25 extra servers for mad Google updates (even seemingly insignificant incremental *push* updates).

    Maybe this “real time web” needs a better definition: is it only related to Twitter feeds and the like? If so, how does that help overall Google search (search “Bird” for info about birds, not tweets)? Will Google be segmented into verticals (search feeds vs/ info vs/ say, shopping)? Is it opt-in for site owners? …so only big, $$ flush, companies (who can afford an extra 25 servers dedicated just to Google) will be in Google top results? etc…etc…etc…inquiring minds want to know. :)

  • “I asked the question of of of Google’s executives”

    can you guys seriously proofread?

  • Too many competitors at hand, WolfRam Alpha and Twitter both are anticipated to emerge as a threat to Google. How well are they going to do? What will happen once Wolf Ram Alpha and Twitter start crawling through the Websites and reputing them alongside? http://bit.ly/NI8QH

  • Too many competitors at hand, WolfRam Alpha and Twitter both are anticipated to emerge as a threat to Google. How well are they going to do? What will happen once Wolf Ram Alpha and Twitter start crawling through the Websites and reputing them alongside? http://buckdat.blogspot.com/2009/05/what-will-happen-once-wolfram-alpha.html

  • I propose that real-time search is in fact too stale and we just need real-time filtering instead. I propose new technology be named a “Chat Room”.

  • Google doesn’t have to acquire another company to provide realtime trends. Some may doubt that Google thought of realtime “results” first. In internet years, Google is ages older than Twitter and as the top search engine it would be surprising if the conversation didn’t come up. Google’s network engineers have access to Google’s answer to Twitter Trending only they call them statistics. If the idea did come up it makes sense that it was shot down because no one has really been clamoring for statistics on what everyone is searching for ‘right now’.

    The power of Twitter’s realtime trending is that its showing you the hot topics of the moment. Safe to assume that the hot topic on Twitter has a corresponding population of Googlers, like myself. I didn’t go to Twitter when the earthquake hit and I live in Los Angeles. My first instinct wasn’t to Blog about the quake or Chat about it, but instead to LEARN about it. So, I Googled the US Geological Service: something that can’t be done on Twitter. There is a vast number of instinctual Googlers like myself who just simply go right to Google and many if not most Twitterers still need more info than the 140 characters they’re getting and go to Google or another search engine.

    Google only has to retool its service to render super realtime statistics on actual search queries. I would add city, county, state, region toggles for good measure. The moment it becomes apparent the number one search query on the West Coast one minute ago was earthquake related voila. They have Latitude and Google maps so if they geo tag the queries they could get real medieval in pinpointing ground zero for individual phenomena.

    To add relevance in a competition sort of way, Google should include tags. Also if my topic isn’t in the top thousand but I want an idea of how many people searched for my topic today, I should be able to get that. Maybe even go back in time and see where foreclosure or broker suing got hot and heavy and be able to tell its fluctuation by the hour/minute. Also if I go to Google while doing a normal search as part of my result I should be able to see where my query falls if I want.

    No they don’t need another company, they just need a lot of redundant hardware, more cooling and bananas for the flying monkies that do the coding over there.

  • Is it just me or is ‘real time’ getting a lot of hype these days?

    I don’t know about anyone else, but my need for knowing what everyone else is talking about ‘right now’ is pretty damned low.

    However, I suppose if you want to harpoon some traffic you could inserting yourself into the ‘real time’ tag stream and show up on searches… yeah, that ads a lot of value for me too.

    Wake me when it gets real.

    • I’m with you. If I needed realtime it would be more out of curiosity. The inherent weakness of realtime is that its even less authoritative as a source than Google results. That’s saying a lot because Google can get you to the page, but you’ve got to vet it.

      Even when there was no internet there was occasional hysteria over a famous person who was supposedly in an accident or dead only they weren’t. Three million people tweeting, chatting or talking about anything doesn’t help when they’re wrong.

      • In my opinion the best case scenario is that real-time news/search and traditional news outlets coexist, but serve different purposes.

        News orgs = investigative reporting, vetted + insightful stories

        “Real-time” = breaking news, hyper-local news, some degree of sensationalism

  • Nice link baitz sirs.

  • Real time search is extremely interesting for news and live stream / feeds.
    I can’t imagine that I get 17,980,093 web sites results, and i wait for 30 second to see that 5 additional sites are added, it is overloaded as it is.
    This is why i got addicted (but i’m not objective) to the Qwiji (www.qwiji.com) search results presentation, instead of trying to decide which result is best for my needs i simply flip through the pages. It’s FUN.

  • I posted a blog entry on: May 18, 2009 @ 22:17
    (Time stamp is 1 hr off, should read @ 23:17)
    http://austincitypermits.com/blog/?p=1370

    Google Web Alert for: “pine street station”
    Mon, May 18, 2009 at 11:28 PM

    That took Google a whole 11 minutes to index my blog post and send an alert.

  • Google wants to do real time ha?! How come my search results for new topics leads to 2005 era sites?

    I think Google is about to die soon… It has become like Microsoft… When I go to google.com, it thinks that I have to go to google.co.in and damn slow too…

  • This doesn’t seem like any ‘amazing’ revelation. Whether for Google or Twitter, real-time search seems so basic as to sorting relevant results in what the application defines as ‘most recent’.

    From a search-engine standpoint, I guess it’s a new way of thinking, but I’ve been observing Google’s calculated movement to integrate social networking elements such as voting links up/down and profiles.

    It seems to me that everyone’s going ape over a hybrid digg/search engine trend. Catalog what’s hot, organize it by date… Am I missing something?

    • “Twitter, real-time search seems so basic as to sorting relevant results in what the application defines as ‘most recent’. ”

      Well, the point is that google does not have the really most recent content put up on the web and google is what everybody is using.

  • Nicolas @quackyduck - May 19th, 2009 at 8:39 pm GMT-1

    Real-time search is only relevant when you’re looking for breaking news. Also great if you’re curious about what’s going on with goldencasinoflow and other spammers on twitter. Awesome.

  • I don’t get the use case for real time search. Seems like hype to me. http://tawk1.com/99

  • I don’t see the use case for real time search. Why would alot of people do it everyday? Seems like hype to me. http://tawk1.com/99

  • The credit goes to Twitter, smart as Larry is.

  • thats the last thing I want google INDEXing my entire site every second, costing me added bandwith…

  • “what is Google going to do about the astounding buzz around realtime search and Twitter? ”

    Astounding buzz wil go away and Twitter will have it’s regular place a popular mini-blog service on the internet. Twitter is NOT a game changer.

  • Real time search ? Who really needs real time data ?
    Just because twitter is doing , does it makes business sense ?

  • Well, twitter search is just searching queries in twitter itself, it doesn’t index other websites, so how come everybody says it’s inspiring Google on real time search? What Google wants to do is to index all web pages and brings back real time results, not even Twitter could to that. What Twitter does just providing search engine for their own content, not for other websites.

  • Since always Google has been in documenting facts of the past: Google Search, Google Maps, Google Street View, Google Earth,…
    That was about storing data.

    Now they need to turn around the company for real-time search which is less about storing data.
    Turning around a large company is not evident only a few successful companies have done this: IBM, Apple, Samsung, …

    Good luck Google.

  • Twitter search not about real-time as much as about the human feed of information: human beings deciding what is most important and where to find the information, that is the beauty of Twitter.

  • Give me a break Larry! You “always thought we needed to index the web every second to allow real time search? Sure you did.

    Maybe before you get to real-time indexing you might want to ensure you serve newer content. Your algorithms weight on links from other sites and therefore older info always comes up first.

    Google has lost its mojo and not innovated anything since search. AdWords – stole from overture/bill gross/idealabs; maps – acquired keyhole, video – acquired YouTube. We are waiting for the next big innovation Google – please show us you’re still alive – please!

    @robblewis

  • twitter sucks

  • twitter search is just searching queries in twitter itself, it doesn’t index other websites, so how come everybody says it’s inspiring Google on real time search? What Google wants to do is to index all web pages and brings back real time results, not even Twitter could to that.i just found this amazing website, it has all kind of businesses from a cleaning companies to estate agents to banks, all I have to do is just type in the business type, and the location and it gives me all the results in that particular area, this website has a huge database of all sorts of businesses, I just joined a driving school which I found on this website. I would recommend this site to all the people out there. We are waiting for the next big innovation Google – please show us you’re still alive – please!

  • All google has to do is incorporate a ’search by date’ and they’ve just done what boilingpage.com and everyone else is trying to do with “real-time search.”

  • Huizhou Universe Hardware & Shelf Co., Ltd. is one of the top designers and manufacturers of various racks, shelves and logistics systems in . Our advanced equipments, experienced engineers and skilled technician and well-trained workers assure us a big design and production capacity.
    Our factory is approved by ISO9001:2000 version, and we get good reputation and trusts from the worldwide markets based on our target of “International quality and sale services satisfy our customers”.
    Our products are ideal for using in factories, airports, army, supermarkets and other logistics distributing center. After years researching and designing, our products ranges includes major assembled and integral shelf series, and our customers are satisfied with excellent service and reliable quality.
    With struggle in the past years, our company is growing very well and becoming strong and more professional, and get positive feedbacks and trusts from foreign customers.

    Our philosophy:
    Sincerely serve our society with high quality products, best reputation;
    And offer our best services to end users with science technology and our experiences in shelf fields;
    Warmly welcome all over the world business dealers contact with us and look forward to our cooperation.

  • Recent studies suggest that young infants, including neonates, prefer speech sounds to nonspeech sounds with similar spectral and temporal properties. ,

  • Let us not settle for the status quo, but instead be visionary, creative and bold. ,

  • True, Google should buy one of these small players and charge against Twitter. Here’s the list of real-time search engines:

    http://www.boilingpage.com/index.php?p=results&n=3001463

    Found this list of real-time search engines in a real-time search engine. How cool is that?

  • Great link plus the real-time search engine http://www.boilingpage.com is frickin’ awesome. Found some great articles about ‘machine learning’ which I’ve been pondering over for a long time and couldn’t find in Google.

    http://boilingpage.com/index.php?search=machine+learning

Leave Comment

Commenting Options

Enter your personal information to the left, or sign in with your Facebook account by clicking the button below.

Alternatively, you can create an avatar that will appear whenever you leave a comment on a Gravatar-enabled blog.

Trackback URL

TC Europe Top 100