Roundup Of Google Blog Search Commentary

Below, a roundup of bloggers and others talking about the new Google Blog Search service that's been launched. For more background on the service, see our Google Launches Industrial Strength Blog Search and Thoughts On & Poking At Google Blog Search posts.

  • Google's new blog search makes a great first impression has Robert Scoble doing an ego search for himself and finding it sucks because his blog isn't tops. Except it is tops but shows "Velveetaland" as the domain. Why? Glitch, hijacking, I dunno. But the actual link does lead directly to his site. He loves the speed and the relevance when doing an advanced search for scoble PDC, sorted by date.

    The only "advanced" thing there really is that he set results to show 100 listings rather than 10. Otherwise, it's just a multiword query. FYI, use the preference options to set results to 100 results permanently. Bad news? That does it for ALL of Google and doesn't let you have a sort by date preference, as I've written. Google Blog Search really needs its own preferences page just for that service, allowing sort by date as a lockdown option.

    More playing with Google blog search has him Robert playing more and liking Google more better, and Testing link search out on Google blog search especially in getting backlinks.
  • Google adds blog search from Dave Winer doesn't seem happy you can get results in either RSS or Atom. Yeah, pick a format and go with it. Of course, if they picked Atom, he wouldn't be happy. So no surprise they decided to offer them both. He's also concerned that if he does a post without a title, Google doesn't seem to "have it." I don't see that. Here are results for his site. Here's the first result:
    46 minutes ago
    NPR has a podcast about the Roberts hearings. I caught a bit of it yesterday and
    it was really interesting. Surprisingly so.
    Scripting News -

    They've got the post -- they just don't have a title because the post doesn't have a title. Solution? Give posts titles. But it may be he's concerned that many of his posts don't appear when compared to Google web search. My other article explains this is the memory problem -- they only have info from March 2005 onward in Google blog search.

  • Google launches blog search – is this the death knell for Technorati, et. al? from Forrester's Charlene Li comes out generally positive about the service. But the idea of Google stripping out "non-blog feeds" like weather and stock quote updates. Not what they tell us. Haven't poked hard enough to test, but if it's a feed, there's a chance it's in there. Only sites they know are in their hand-picked source list for Google News are potentially removed from Google Blog Search, we've been told.
  • Welcome to the Blogosphere, Google! from Technorati's David Sifry doesn't have him putting Technorati in a casket but instead saying the competition will make things better for everyone.
  • Google Blog Search is live from Technorati's Niall Kennedy talks about blog search as a potential new testing ground for search marketers. Not really. It's based off feeds, and the ranking is going to be much different than the full text indexing of web pages that web search uses. But then again, if you haven't yet starting running a proper blog, time to start doing them. As I said at an SES conference recently, they're the "acceptable" mirror site. IE, have a regular site and a blog, and that's just fine to the search world, assuming they do different things. And you should, because as you can see, to play in blog search, you need a blog. An ordinary site doesn't get in to have fun.
  • Google Blog Search reviewed at Blog Herald thinks the link counts suck compared to the competition. Chances are, this is because the Google blog "memory" stretches back only to March 2005, as I wrote. Similar criticism over size of coverage, and chances area again that history is the issue.
  • Google launches Blog Search at Six Apart, spotted via John Battelle, focuses more on Six Apart's AtomStream service that's in development, that flows out all the content of LiveJournal and TypePad content to any tool that wants them. Pinging and crawling might be eliminated if everyone just flowed out full information this way.

    Yes, except that blog/feed search services are increasingly encountering the same spam problems that web search has had to grapple with. Expect that if anyone can flow content, some are going to making use of this nice cloaking mechanism to be misleading.

    THAT's why web search engines crawl. The idea of flowing information to them isn't something they never thought off until now. They've simply learned not to trust what people might give to them. As blog spam continues to grow -- IE, "fake" doorway page style blogs -- the blog search world is going to learn in short order that flowing info may not work as well as they think.

    I really do hope some better solution will come about so people can feed content into the major services, rather than just sending URLs. Certainly Google's Froogle has seemed successful on this front.
  • Google Launches Blogsearch from InsideGoogle posts thoughts here and finds that overall, it's likely to pull people from other blog search services.
  • Google Blog Search Launched is discussion at Threadwatch.
  • Via Barry's Search Engine Roundtable site, see also discussion at DigitalPoint Forums, Cre8asite Forums and WebmasterWorld.

  • Google Blog Search Launched is discussion on our Search Engine Watch Forums.


  • Google Launches a Blog Search from Tara at ResearchBuzz who likes somethings, hates others, but especially thinks the default should be by date. I'm with you, Tara -- sort of. Problem for Google is that when you search by data, relevancy drops like made. Searches for [google] still bring back tons of junk/spam. Tara also details some of the URL switches used to get the number of results back, for those who want to dig really deep. She says site: works; I still find it does NOT.

    WAIT, MYSTERY SOLVED! I REALIZE IT WORKS IN COMBINATION WITH ANOTHER QUERY. In other words, brings back nothing but mary hodder will work. Google web search used to not allow site: as a standalone search but that was changed a few months to a year ago, if I recall correctly. Blog search has yet to realize this and depends instead on blogurl: as the command.

  • Dear Google, Thank you for blog search has Jason Calacanis expecting blog search to show up on the Google home page. Yeah, maybe. More likely, I think you'll see it combined with news, especially since there's only so many buttons you can put up there. Note to Jason -- no search engine wants to look like this, and people don't see those tabs/links anyway as the article that link leads to explains.

    Jason also completely overlooks the fact that plenty of news search services already put great small and individual publishers into the results there alongside MSM -- mainstream media or "traditional media." In fact, I'd wager the vast majority of news search sources in any major news search service are NOT traditional media outlets. So ease off the idea that this is a precursor to the big leveling of web society. Web search was already a leveler and remains so.

    Realistically, news and blogs go together as well as news and opinion go together. The two sides of the same coin and all that. Need that illustrated more. The new Memeorandum site does the blend well, as I explained yesterday.

  • Google Launches Blog Search from Mary Hodder who has done a great series on blog search recently finds an ego search for her name doesn't get all references (it's likely the memory problem mentioned above, Mary).

    She also doesn't like that a serach for her own blog name bringing up her own pages over and over is annoying. Yep, and if if were web search, Google's clustering (showing you only up to two pages from the same site per results page) would have solved this. No clustering for blog search. Easy workaround, however. napsterization There you go -- your self-referential problem disposed of.

    She also notes no change in relevance when doing a link lookup and sorting by date or relevance. Betcha I know why. Google may simply not want to tell you what link it thinks has the most link juice flowing to a site.

    Overall, she thinks they've got a lot to learn about blog search. Probably, but it also sounds like that perhaps tailoring some of her regular searches with a few search commands might also get more of what she wants.


  • One Service, Two Faces from Ken Yarmosh highlights how the Blogger version makes a search within and find all posts feature available for listings. You could do this with the Google version, just not as easily.

  • Google Introduces Blog Search from Tristan Louis wants to see one of the major search engines offer their own ping server. From talking with Google, they seem more behind the idea of supporting the development of FeedMesh as a solution.

  • Blog Search: Redirects and Indexing from the official Blogger blog, spotted via InsideGoogle, notes that the redirect page you see when clicking on a result in blog search will be going away, now that it is no longer needed to help protect the secrecy of the project.

  • After Just a Few Hours With Google Blog Search: Comments and Wishes from our own Gary Price has him giving a laundry list of wish, as well as showing how non-blog content is in the blog search engine and how SafeSearch fails to keep you safe.
We may add more items as we see them and as time allows. In the meantime, need more? See the new Memeorandum Tech page, and the new Google blog search is the top item, with tons and tons of links to discussion from across the web.

Want to comment or discuss? Visit the Google Blog Search Launched thread in our Search Engine Watch Forums.