They have inferior search technology on several fronts (Ask generally provides better quality responses for general Web search).

They only control about 35% of REAL Market share (as measured by number of visitors per search engine and number of sessions per search engine).

They are getting their butts kicked in major non-American markets like Russia and China.

And they aren’t called The Search Engine That Spam Built (TSETSB) for nothing. It’s easier to get Web spam into Google than into the other search engines.

People need to stop fussing over the estimated number of queries performed at search engines. If you have to run 3 times as many queries at Google as at Microsoft, that’s a pretty solid indicator that Google isn’t showing you the best results.

In terms of profiling, i feel that to provide the ‘best’ answer to all search queries (or just questions) the phrasing of the query is insufficient, you mush also have context – geographic, other recent questions, personal preferences, near future events – as it is inefficient for the query to include all these factors
(eg if i asked a friend where is good to eat nearby, they may take into account that i can drive & have my car available)
It may be undesireable, but will tend to improve the results, even if that’s only by eliminating irrelevancies

nmw: That didn’t sound correct when I first read it. Half a century?

Then I went back and did the math. Eugene Garfield, Science Citation Index (SCI). Where the idea was to index papers based on citation “hyperlinks”, and use that index for improved information retrieval, ala PageRank. Check out this quote from a 1955 article by Eugene Garfield (http://garfield.library.upenn.edu/papers/science1955.pdf). I’ve bolded those passages that essentially illustrate the core Google PageRank idea:

“Under each code number, for example, 3001-6789, there would be listed other code numbers representing articles that had referred to the article in question, together with an indication of whether the citing source was an original article, review, abstract, review article, patent, or translation, and so forth. In effect, the system would provide a complete listing, for the publications covered, of all the original articles that had referred to the article in question. This would clearly be particularly useful in historical research, when one is trying to evaluate the significance of a particular work and its impact on the literature and thinking of the period. Such an “impact factor” may be much more indicative than an absolute count of the number of a scientist’s publications, which was used by Lehman (3) and Dennis (4). The “impact factor” is similar to the quantitative measure obtained by Gross (5) in evaluating the relative importance of scientific journals, a method later criticized by Brodman (6) but used again by Fussler (7).“

So Google in 1998, Garfield in 1955. That’s 43 years. Almost about half a century! You’re absolutley right. Amazing.

With user education, not only will (as JuliaK says above) competition over the quality of the search RESULT be put back into user-beneficial play, but the PROCESS will be up for grabs, too.

The problem is that Google has conditioned us to believe that one process is the only available/best process.

Imagine what would happen if many Americans started believing that a single source (process) of news was the only source of news. Let’s say, Fox, for example.

Loss of diversity is a bad thing.

It will take a new ‘something’ that won’t topple Google in the search market but in another field not yet invented or brought to the masses.

Just my 2 cents.

Thanks,

- Holger