Google Killers. I’ve mentioned them before (here, for example). Many of these so-called Google killers are akin to child stars from the 1980s: Where are they now?
Now we have another “Google killer.” This one is called PowerSet and they have very clearly set their sites on Google. In an article on MSNBC, PowerSet claims to have developed the most power natural language search engine.
From the MSNBC article: The distinction means Web surfers will theoretically be able to get more meaningful results by typing more precise search requests in the form of straightforward questions like “What did Steve Jobs say about Apple?” instead of entering an ungrammatical mishmash like “Apple Steve Jobs said.”
While it is true that keyword searches are not always a shining example of the Old Kings English, the fact remains that people are now accustomed to searching in a specific way. Search is now an established industry, more than a decade old, and the single or double keyword based search habit is well established by searchers all over the world.
Assuming PowerSet is correct – that they have indeed built the most amazing natural search algorithm – they still have to pretty much retrain the already instilled habit of a searcher.
Lest we forget, Ask.com has gone down this path and stepped away from it. Ask.com started as AskJeeves.com and was marketed as the natural language search engine. The idea behind the experience was a searcher would type in a fully formed question to Jeeves. But a few years ago, the butler went away and the Teoma backbone took over to better serve the searcher.
But let’s also ask the question: is the assumed lack of natural language searches a problem? I decided to conduct a small test to see. I conducted a variety of searches, all in natural language and all on Google.
“What do tadpoles eat?”
That came to mind because my son asked me that question a few weeks ago. I found my answer very quickly, in the first few results. (See the results to the search here.)
“How many mammals are there?”
(Again, a question my son asked me. See where I’m getting my inspiration?) Again, the answer to the question is found in the first few results on the first page. (See the results to the search here.)
“How big is the universe?”
(Guess on how I came up with this one??) Again, there is no lack of directly relational
results on the very first search results page. (See the results to the search here.)
“When did Steve Jobs announce the new Ipod?”
(OK, I thought of that one on my own.) But again, the results on the first page of
Google clearly lead me to a direct answer. (See the results to the search here.)
Clearly, in my small test, Google has no trouble with natural language searches at all. This indicates to me that overall, Google’s ability to handle natural language searches is not a problem if someone should happen to search that way.
I also generally have issues with anyone that calls themselves a Google killer. Maybe they do it
for the press or maybe they do it to get Google’s attention. But if a lofty claim requires lofty results. It starts to sound a lot like the guy in the neighborhood that is good at basketball and claims he could play in the NBA if he wanted to. Sure you can, buddy.
After all, there are plenty of companies that have made that very claim that never came close to delivering.
Bonus: In order to find the picture of Gary Coleman, again, I used a natural language search on Google. “Where can I find a picture of Gary Coleman?” Obviously, I had no trouble.
{ 1 comment… read it below or add one }
I like watching these “Google Killers” develop. Some actually have some interesting technology but I think you’re right that they are climbing in the ring with a huge giant. From my personal observations, I think Google is working on all these “semantic web” or “web 3.0″ type technologies. Before Universal Search was announced by Google there were evidence of it. Also, it seems that Google is trying to “predict” what people are looking for. If a search for “tvr tuscan” is done, the SERPS are divided into 3 sections. I have run across this a few times.
You must log in to post a comment.