There has been a significant uptick in the number of press releases in regards to new search engines trying to accomplish the next phase in search: natural language search.
John Battelle, in his book "The Search" not only touched on it in the last few chapters but pretty much put out the claim that the company who could best determine relevancy will win in the hyper-competitive world of search. Two examples that initially came to mind are the simple words "dog" and "pepper." There is more than 1 meaning to the word dog, but besides the obivous one of a four-legged mammal, I have to assume you have heard of the phrase "that was a dog" where that could refer to maybe a slow runner/horse in a race. The meaning of dog in this scenario is slow, or weak, or basically something that is inferior, which is quite different from a golden retriever, poodle, or an Australian blue shepherd. A second example that came to mind is "pepper" -- pepper definitely has more than 1 meaning, as it can be the spice you add to food, but it can also be the vegetable (both green peppers and red peppers) that so many people in this world love to consume, myself not included in that group. If you are searching for pepper, how is an engine supposed to know which pepper you are looking for - the spice or the vegetable? Well, Battelle referenced the fact that search engines are getting much smarter and in their SERP's (search engine results pages) display the accurate one based on your past search query history. That is one way that search engines are trying to become much more relevant and how they are trying to improve its users' experiences.
Having said that, serving up search results based on past search history is only one part of the process. What these new companies creeping up on the web are trying to overcome and produce are natural language search results. What does that mean? My best definition is a search engine will be able to answer the question you are asking based on the context of your query. The one million dollar question, or maybe it's more like the one billion dollar question, is will that allow search as a distribution medium to continue its amazing growth? Also, will these new companies truly provide a better search experience?
I will touch on a couple of the companies that I have been reading about and generally say that I will wait in angst as some are still in "stealth" mode (a term which I thought was going to die after the Internet's dot-com bust 7 years ago but apparently is back in vogue). Powerset has been in stealth mode since last November but has a very impressive list of backers and management team. When someone says Internet advertising is not rocket science, they are correct, but the CEO of Powerset - Barney Pell - truly is a rocket scientist. After working as a VP of Business Development for a company called Stockmaster.com, Pell went back to work for NASA and Mayfield Ventures.
Hakia is another search engine that is gaining momentum and one that I have not used very much, which very well might change. I typed in a simple question "how are hurricanes created" and the first organic listing - http://hakia.com/search.aspx?q=how+are+hurricanes+created - pointed to an article that appears to answer my question. I then typed the same question into Google and Brainboost (an Answers.com company). The first organic search listing on Google, for better or worse, had the same kids.earth.nasa page that Hakia pointed to - http://www.google.com/search?hl=en&q=how+are+hurricanes+created&btnG=Google+Search
Brainboost, on the other hand, served up a results page that pointed to a couple of blog postings from journals.aol.com and huffingtonpost - http://brainboost.com/search.asp?Q=how+are+hurricans+created&Submit=ask#
If I had the luxury of being able to do a couple of examples across each engine, I'm not sure if I would get similar results to my hurricane query, but at the very worst it's informative and I can say I learned something today.
On sort of the opposite side of the fence from predictive analysis and natural search algorithms, a couple of companies are trying to create search engines whereby human editors will determine what sites are the most relevant to what people are looking for. The first company - ChaCha.com - has been featured in several newspapers very prominently, and it will be very interesting indeed to see if they can steal some market share that Google and Yahoo continue to dominate. The second company - cranky.com - is a division of Eons, Inc., and has a good chance of succeeding as the population of the baby-boomers continues rising daily. Cranky.com is focused on users that are 50+ and many of those people (my parents included) are slowly but surely understanding search engines but are frustrated by the sheer volume of search results that many times lead them astray and different directions.
I really wanted to cover a couple of vertical niche search engines in this article as well, but I will either write about that in a couple of weeks back here OR on the corporate site of www.motiveinteractive.com
Scott Pannier | Director of Media
Motive Interactive | scott@motiveinteractive.com
http://degrees.educationsearch.net
http://nextgentele.com
