A friend just sent me a link to Google Blog Search, and I spent some time looking at it. My first reaction: its about time that Google added in Blog Search. They have been building product after product, and ignoring something that is their core competency.
In the meantime, too many results from blogs often made web search difficult to wade through. There are two search situations – sometimes I want to search the web, find the most relevant item from results all over the web. At such times, I don’t want blog results – what some blogger said, just because its recent.
At other times, I want to find recent conversation about the topic. It should be relevant, but recency is more important. At such times, I want blog results.
In neither case could Google get me what I wanted – for a web search, the results were too influenced by blogs and recency. For a blog search, it was not sensitive enough to recency.
Which is why its great that at last Google has introduced “Blog Search“.
Some initial impressions (after a very brief look).
a) The interface is clean, and works as one could expect. I am glad I no longer have to deal with Technorati’s cuteness.
b) Its fast.
c) Its in Beta, which makes sense, unlike other Google products, such as Google News, which are in perpetual Beta.
d) It does not seem to be using Tags, or Categories in any fashion – this is not unexpected, coming from Google. (As a note, this could be what Technorati ends up focusing on, and building their strength around).
(Interestingly, if you search for Technorati on Google Blog Search, then you find a bunch of posts with the words “Technorati Tags”.)
e) I don’t get how it is operationalizing relevance. My reaction is based on a cursory look though. When I type in “Rashmi” – for the first page results, I get a bunch of postings from my blog and for a few other Rashmi’s. The next few pages show more postings from my blog – but the order in which they are presented is entirely opaque to me. (If you understand how Google is ordering by relevancy for Blog Search, please post a comment).
f) “Sort by date” is transparent – simply ordered by time. (Note: I see lots of Blogger results in this view – is it biased towards Blogger?)
g) Currently, in order to access Blog Search, you need to either go to the url directly, or to the “More” link on the Google Home page. I think that “Blog Search” needs to be a front page link, along with “Froogle”, “Images” etc. Personally, I will be using it more than any of the other links, so I am hoping that Google will add it in.
h) Apart from links to posts, it also gives you a link to “Related Weblogs” at the top of the results. Currently, it seems to present only one weblog link.
i) Interestingly, it considers Flickr a blog, as someone pointed out on Metafilter. On a related note, Niall Kennedy points out what Google is doing is RSS search, not blog search (Google is not parsing the HTML of the blog, it is using the feeds to break up the content into title, entry etc. The Google indexer indexes the page as a whole, so this makes sense from Google’s perspective – to use the RSS, rather than design a new indexer. Also, as RSS becomes more popular, there will be all sorts of other content available through RSS – so this is a more flexible approach as well).
j) I cannot help wondering what a “I’m feeling lucky” button would look like for Blog Search. That could be an easy way to find the most well known weblog for a particular query, especially a name. For example if I typed “Dave Winer”, and clicked “I’m feeling lucky”, I would be quite happy to be directly taken to his blog. I never use the “I’m feeling lucky” on the Web search, but I think it might be a natural fit for name searches on Blog Search.
(I never thought the day would come that I would want to use the “I’m Feeling Lucky” button!)