Wow, what a day.
Due to some meetings I was only able to hit two sessions.
The first one I went to was “Social Search Overview” – a topic until recently I have not known much about. Well, I soon will, because according to all the speakers, it will be the “next big thing in search”.
Moderated by Chris Sherman, the first speaker was Grant Ryan from Eurekster. He was extremely jet lagged but told a really good story abut how you can create your own niche search engine. Some of the things he talked about were that 30-35% of queries are subjective in nature. The example he used was “Cool Lamps” – what makes a lamp cool? Well, with social search you can put your reputation where your mouth (or ego is) and claim it is a cool lamp.
What is social search? Perhaps the best definition I have heard is that it is “the online water cooler which has a buzz cloud surrounding it”.
Tim Mayer from Yahoo! spoke next. Tim is very much an algorithmic man but also spoke about how social search will assist in solving the search problem, through tagging and trust. He shared some of Yahoo!’s offerings in this space.
Nils Pohlman from Microsoft spoke next, with a similar story to Tim.
Essentially social search will be driving the next wave of search and the two followers (Yahoo! and MSN) are seeing it as their way to catch up to Google.
I went into this session all excited and hoping to see some new innovations. I was disappointed. Peter Norvig from Google spoke first about how many of the Google products started with no master plan. They were simply designed from the bottom up and ran as projects to see where they went.
For example, G-Mail was the creator of AdSense. A Google engineer was struggling to find all his mail and wanted to store it online. The concept for G-Mail was born but he wasn’t sure how to pay for the storage. He suggested showing adverts on the page. The need to make them relevant meant that AdSense was born.
G-Mail launched well after AdSense for product reasons and not technology reasons.
The other thing that Peter showed us was a tool that can recognize faces in a photo. It doesn’t work out who the person is, but with a fairly high degree of accuracy, will find faces in a photo. It will also determine with about 95% accuracy if the face is female or male. This technology will probably go into Picasa first and then into Google image search at a later date.
Yahoo!’s Bradley Horowly spent a lot of time going through the Social Search side with not a lot of new stuff. He did talk about “zone tagging” which is where you take a picture with your cell phone, post it to flickr and have the meta data tagged. It could be tagged either by the user at the time of uploading or by the cell phone’s location through the Cell ID.
James Colhorn from MSN talked all about AdLab – the lab part of Microsoft’s AdCentre. Rather than tell you about what is on offer, visit the AdLab yourself at adlab.microsoft.com. In true Microsoft fashion he labled it as a “cool tool”.
Hopefully I’ll be posting some photos tomorrow after the Google Dance.