Tim Mayer of Yahoo! Search wrote in the Yahoo! Search Blog that in search engine results it is “interesting to note how little overlap there is – nearly 85% of first page results are unique to one engine” and that it “sort of makes you wonder what you are missing if you are stuck in a search engine rut…”
It is true that specific pages listed in search engines are different, but they are all essentially the same in terms of what they offer the user. In fact there is very little variation in search engine results. It is difficult to find useful information about a range of topics, because getting ranked well is mostly an investment by the site owner.
Of the major search engines, Google continues to be most popular, and it is no wonder why. Google is fast, has an extensive index, and is an excellent marketer of its services. But as far as I’m concerned, MSN and Yahoo! provide results that are almost always just as relevant. MSN is slow, and often produces search errors. Yahoo! doesn’t index sites nearly as fast as or to the same extent as Google. Yahoo! doesn’t provide the best results for searches on sites that are rare and new, but Google doesn’t either, because it implements an age delay when ranking new sites.
All the search engines use the same basic criteria to rank sites. They differ in the weighting they give this criteria, but there’s nothing novel about search criteria. There’s certainly nothing novel about Google’s PageRank; MSN and Yahoo! also use linking criteria to list results. If a site is ranked well in any two of the three major search engines, it shouldn’t be difficult for it to get ranked well in the third (if it isn’t already). Search engines also use the same basic display: ads on the top and sides, and something around ten “natural” search results (also should be considered ads for the most part) in the middle.
Search engines are (unsurprisingly) uniform.
Tim Mayer Is Wrong
If Tim Mayer wants users to get out of their “search engine rut,” he should focus on making an original search engine with a solid ranking algorithm – to truly provide a gateway to all the content the web offers.
How Search Engines Could Improve
Returning basically the same ten first page results for months on end is ridiculous considering the wealth of information the web offers. A good search engine would identify general queries that are relevant to millions of pages, and try to deliver variation in results on a regular basis.
Most of the web remains invisible to search engines. A good search engine would make greater efforts to index the invisible web.
A good search engine would distinguish between commercial sites and non-commercial sites, delivering non-commercial results for queries that seem non-commercial in nature. Commercial sites are interested in owning all queries relevant to their industry, and it results in phony white-papers, misleading forums, and a general suppression of good, useful information.
A good search engine wouldn’t have to keep its ranking algorithm secret, because a good ranking algorithm would be impervious to manipulation. In fact, search engines would want to reveal all the details of a good ranking algorithm because it would encourage webmasters to produce better sites.
What good is going to different search engines if all they offer is sites with different domain names, all providing basically the same content? How could a search engine look at the content of prospective results and determine if the meaning is all the same? Why does it seem like search engines are ignoring these questions?
