April 26, 2008

The Search

It is interesting to see the plethora of search engines coming up almost every month trying to solve a particular aspect(problem) in searches and claiming them to be potential 'Google-Killers'. I admire Google for its ability to 'create' consumer markets than just following or doing what the consumer needs. This is very important for a product firm which does wish to succeed and has the guts to run the extra mile and implement some bleeding edge technologies. Google took a very simple requirement of enabling better data management on the web to a completely different level and is unarguably the best in its class now. Numerous books/articles/blogs have been written and eulogies sung in praise of Google; but at the same time I would like to raise a concern that I have been noticing for the past couple of months.

Google's relevance of results is decreasing day-by-day. I see loads of spam in Google's BlogSearch. There is no effective 'innovation' in the data-visualization and the way the results are being presented to the user continues to remain the same; and the Ads are also being presented in the same format.

IMHO, i would say the following 3 features to be 'most-essential' for the next search-engine startup to be successful:
- Duplicate Elimination (Chris Manning's group in Stan-NLP is doing some amazing research in this, and i guess a stealth-mode startup would have already started)
- Social searches (Something similar to what the Israeli startup Delver is doing - but of a much greater magnitude and more 'intelligent')
- Data Visualization

Update: Got 2 new books to read:
1) Phantoms in the Mind , By V.S.Ramachandran and Sandra Blakeslee
2) Indian Unbound , By Gurucharan Das