We’re really pleased with our launch – great exposure, lots of users playing around with the site, many kudos along with some constructive inputs. It was a good week, generating nearly a million page views, which really blows us away. It also underscored how helpful bloggers are about getting the word out – if you’re one of the folks who blogged about us, thanks – very much appreciated.
One data point that really jumps out is the use of our Sphere It! bookmarklet (we’ve had tens of thousands of downloads) and the high quality comments it received in reviews and personal emails. It’s our favorite app too so we’re pleased that others have quickly started using it to look for related blog content to the content you’re reading.
A few questions came up about the difference between Sphere It! and Technorati This! Both products try to give you content from the blogosphere that’s related to the web page you are currently reading. Sphere It! does this by doing semantic analysis of the page you are on and finding blog posts that talk about the same topic. Technorati This! does it by looking for blog posts that link to the page you are on. So if you are reading an article on the new CIA director, Sphere It! will give you the most relevant blog posts that also talk about him. Technorati This! will only give you the posts that happen to link to the specific article you are reading (which often miss or have no results at all). Here is an example (you need to download the Sphere It! and Technorati This! bookmarklets) that helps illustrate the point:
Go To Time – here is the current lead article on their home page (today at 2:37 pm) called “What’s Gone Wrong for Britain’s Tony Blair”.
Click on your Sphere It! button and it should bring you to a sphere results page – I’m showing 1,000+ blog post results talking about concepts covered in the article.
Now click on Technorati This! and you get 4 blog post results which are all duplicates.
This above example exposes some weaknesses that a link based approach (Technorati This!) has vs. a semantic analysis approach (Sphere It!):
- A lot of content has zero links from bloggers therefore the results set will be zero or shallow
- The freshest content is unlikely to have any matches or very few matches because bloggers haven’t had enough time to link to the content
- When there are results, they’re time based so you don’t get relevance, just the luck of chronology
Sphere It! results are:
- Immediate – we’re not dependent on links – we just look at the content, figure out what it’s talking about and then look at blog content and see who’s talking about similar things but from a blogger perspective
- We’re showing results for contextually relevant as well as links to the article