Who will win the new search engine war?

Which algo wins your vote?

[Liane]

I've been wondering why, in mid November (at the busiest time of the year no less), Google chose to take a flying leap by rolling out an entirely new (and very frightening) algorithm. Of course we now understand (three months later) that it is a semantics based algo which includes stemming. At the time though nobody was able to make head nor tails of it because, quite frankly, the results gave no indication of what Google was trying to achieve other than search engine suicide. (Sorry GG)

(Of course, they cleverly did all of this while Yahoo was still using their index. Any market share lost went to the also rans and should be easy to win back in a short period of time. :) )

Clearly, the results after Florida were horrendous. If the powers that be at Google had been as happy as they professed to be at the time, major alterations to the algo wouldn't have taken place since then. And who knows if Google's results will stabilize for any length of time once the results from 64.233.161.99 have gone live?

Assuming there will be a period of stabilization, we can further assume that the past three months were an effort to have the new algo operational in time for the rollout of the new Yahoo search technology ... which is not bad at all.

Its still too early to tell, but its interesting to speculate who will win in this head to head battle for market share. A week ago, I would have bet on Yahoo as it didn't appear that Google was even close to delivering reasonable and relevant results which didn't include hundreds of directory listings in place of the meat and potato sites. Today, I would put my money on Google once again ... provided the 64.233.161.99 results stick or become even more relevant with time.

Semantic search or "themeing" is certainly the wave of the future as it makes sense to deliver pages/sites based on relevance to the individual search. However, I believe stemming will ultimately put Google leagues ahead of the others in delivering the "most" relevant results.

I did a few searches over the weekend using 64.233.161.99 for a few things I am looking to buy in the near future. I found exactly what I was looking for on the first page of the results and have book marked those pages lest the algo change yet again. ;)

I did those same searches on Google last week and found nothing but directory sites, ebay and amazon listings, got frustrated and went to Alta Vista. I found what I wanted on AV ... which is saying a lot for AV!

I did the same searches using the new Yahoo search and although I found "some" relevant sites, I didn't actually find what I was looking for.

I live on a small, remote island in the Caribbean where a large majority of consumer products simply aren't available. In the past, we relied 100% on magazines and catalogues as our source of information. Every month, ships came in loaded to the gills with individual items purchased through these sources.

Today, many of us use the net to do our shopping although magazine and catalogue shopping is still very important. The ships still come in, but there is far less merchandise on board but the Fedex and Post Offices are always jumping!

Rather than judge the search results based on how our own sites rank, try to find something you really want to buy on all of the search engines and compare only the first page results. Did you find what you were looking for in 10 clicks or less? In my estimation ... that is the only criteria I need to convince me which is "the best search engine" out there. If I can't do that, I get frustrated and tend to revert back to my old buying habits ... using magazines or catalogues to help me find what I am looking for.

How people search is an interesting problem to solve. I have always blamed myself for my inability to find stuff on the net. I am not a very sophisticated searcher and just assumed I was inept.

If you were to assume that you live on a remote island with no department stores or specialty shops offering many items for the home, then go try to find a light weight, 100% cotton blanket in hunter green ... or 100% cotton pillow protectors with at least a 400 thread count for goose down pillows.

The new Google algo is (IMHO) the best yet. I'd be interested to hear what other people discover in their quest for any information or product you are looking to find. Can you find "exactly" what you want (without knowing a brand name) in ten clicks or less?

[superpower]

I think the new algo is good, but I think it could be much better. There are still a huge amount of problems.

I think one algo which satisifes everybody, or even the vast majority of people, is probably just wishful thinking along the lines of the holy grail.

As other searches improve they will be seen as viable alternatives and used more frequently (to the detriment of Google). We are witnessing Google's peak. They may continue to rise for a while, but I doubt a few years from now they will be as monolithic as they are now!

Liane, what island are you on? How is your connectivity? I'd love to live an island too for a while :)