DISQUS

louisgray.com: http://blog.louisgray.com/2009/03/technoratis-revenge-site-is-beating.html

  • jhylton · 9 months ago
    Louis, Google BlogSearch lets you search on links or regular keywords.
    Our link operator is comparable to the Technorati reactions search. I
    tried the query <a
    href="http://blogsearch.google.com/blogsearch?q=link:www.louisgray.com">[link:www.louisgray.com].
    It returns a few results that I couldn't find on Technorati, and the
    three posts you mentioned that you could only find on Technorati.

    Unfortunately, half of the results of the first 10 results are coming
    from links in blogrolls. We're working on that problem and hope to
    have some improvements released in the next couple of weeks.
    Technorati isn't immune from this problem. It returns
    http://www.makingworkflow.com, which has a link in the blogroll.

    You can use the link: operator will an arbitrary url or url prefix. So you can track people linking to FriendFeed or Twitter.

    The keyword search will find a slightly different set of results. I
    found a post that copied a bunch of tweets, and one of the tweets
    included the url of your site (but not a link). The ranking of these
    results, even in sort-by-date mode, is a little different than the
    ranking of link queries. We try to filter out low quality sites and
    remove duplicates, among other things.
  • Louis Gray · 9 months ago
    J, thanks for the detailed response. I expect you guys will have this all solved relatively quick. That's why in the rare cases another service does an end around, and comes out on top, it's newsworthy. Looking forward to the updates.
  • tonio · 9 months ago
    Google "... maybe too good. The company's over-aggressive spiders ... "'

    yeah like when they pound one url 100 000 times in one hour!
    Note it's the bear's paw coming down, Google just took a nosedive today after the CEO sneezed, or maybe Louisgray sneezed, or maybe Technorati is doom to Google based on "David X. Li"'s formula. Who the f. knows.
  • charlieanzman · 9 months ago
    Louis - Might have a little to do with their new crawler ... http://technorati.com/weblog/2009/03/479.html Was surprised at a variety of results there lately too but they have a long way to go.
  • craig · 9 months ago
    Seems like most blog search engines are pretty shoddy and don't have an absolute way to define what they are looking for.
  • jalichandra · 9 months ago
    Louis, very insightful post. Most people expect Technorati search to be 'just like Google' when it's not, nor was ever intended to be. So it's nice to see someone recognize the difference in approach, and unique results Technorati delivers.
  • Louis Gray · 9 months ago
    Richard, thanks for your comments. Technorati is one of those sites that has so much potential, and most people aren't sure what it's trying to be these days. I would love to meet up with you and the team and learn more so I can be a better advocate, or you can use us for informal QA and ideas. Thanks for checking in and in your blog post following up.
  • AndyBeard · 9 months ago
    Google's currently problem is that they might be attempting to be more like Technorati

    In the early days they used to index RSS, now they have content being crawled, and the data is used in a multiple of ways, otherwise they would end up crawling and storing the data multiple times.

    I did ping Matt Cutts about the blogroll problems and I believe he fed the info through to whom it might concern, which is never clear.
    That last point is probably my main gripe with Google Blog Search compared to Technorati - Google doesn't really have a public face for Blogsearch
  • Matt Cutts · 9 months ago
    Hey Andy, Jeremy Hylton is a blogsearch person who has been doing more posting outside of Google.

    I think part of this issue is that the people that are complaining are doing link: queries rather than normal searches over blogs. In general, the change from indexing only content in feeds (which can be partial and incomplete) to indexing the full blog page via crawling is a good change for the vast majority of blog searches. It can lead to additional matches in link: searches that people don't want because of blogrolls. The blogsearch folks have done a good job of reducing blogroll matches, but there's a lot of unusual blogs out there that do blogrolls in strange/nonstandard ways.

    Some people dislike the additional blogroll matches (esp. when they only want new links), but overall the change to index the full content of a post is better in my opinion, because it gives searchers a much more comprehensive index.
  • AndyBeard · 9 months ago
    Thanks Matt, poor Jeremy is now being stalked on Friendfeed and Twitter :)

    I wolld agree that the primary ranking algos are much better than what Blogsearch was using a year ago... heavily weighted to title tags, though occasionally that also it also backfires when fresh content is pulled into the primary serp that has little or no content... just the word in the title.
  • Louis Gray · 9 months ago
    Andy, I do know Matt mentioned him looking into it a while back on FriendFeed. What Blog Search hasn't done is made a specific way to look for linkage, or "blog reactions". If they did, it's possible Google would have the largest, most accurate, index.
  • Eight Women Dream · 4 months ago
    Nothing ever worked for me with any of my blogs on Technorati and I think the game has changed with the entrance of Twitter and Facebook. Why use them when I can control my own marketing through Facebook and Twitter? Not to mention how bad their customer service is and that they rely on Alexa for screen prints of websites ... as if Alexa matters anymore either. Last I heard Technorati was under 30 employees and going fast.