se algo's

holy grail algo - myth or reality?

[rayjam]

Is it possible that the some se's are using a different algo according to the search term. I ask because of the varying stats from researching different search results. The quest to find the ultiamte combination of elements which would work across all search terms could be a fallacy. And in fact what works with one search term will not work with another - anyone have any thoughts on this?

[seth_wilde]

There are some minor differences on some engines, like AV using exact matching for popular phrases. But I think whats usually mistaken for seperate algos is actually different levels of competitiveness.

If you analyze the results of a relatively uncompetitive phrase you could probably rank very well just by tweaking keyphrase weights. If you then take this knowledge and try these same weights on a competitive phrase, your results are not going to be nearly as good.

Beacause the first set of results was uncompetitive, the SEO/webmaster that designed those pages didn't need to match the algo as closely. With the second set, the pages were probably tweaked to perfection and had lots of good incoming links. So it's not really seperate algos, its just that some serps are closer to the actual algo than others.

[rayjam]

Has any research that you know of been done on the incoming links factor? I think one of the most interesting starting points would be to have a list of the elements which decide rankings in order of priority like for example - inktomi = 1. title 2. metadesc 3. alt tags etc etc. Do you know of any results of this nature for any of the SE's.

[seth_wilde]

"Has any research that you know of been done on the incoming links factor?"

There's all kinds of white papers about different forms of link popularity...

"a list of the elements"

I myself haven't done any research to seperate all the factors in a numerical order. Generally the title, text, and links are going to play the biggest role in your listing.