truthbetold - 1:42 am on Mar 28, 2004 (gmt 0)

I understand how PFI works on the backend.

The most important point that MakeMeTop made here is that PFI gives a clear boost due to the fact that it must ESTIMATE the link pop of the page rather than using the normal link structure data of the web. This is how they can still say "we just apply the regular algorithm on PFI pages" and yet still give a boost in reality so they make more money.

This is especially true for the XML Feed programs which their partners are feeding Yahoo. Since they're not real pages, but are rather just fake auto-generated pages which can be manipulated, there is no regular link data to associate with it from the regular web. Therefore, they just estimate the link pop factor for the auto-generated page... which ends up giving it a boost.

Im very happy to hear that Ask Jeeves... and now MSN have both recently announced that the trusted feed PFI programs hurt relevancy and cause mistrust of the search results and have decided not to use them in their indexes (MSN said they wont use PFI in their new search engine). I'm confident that Yahoo will soon follow suit and will dumpt their new PFI program. They have to be smart enough to know that if they want to compete, they MUST have relevant search results that people can trust and PFI simply does not jive with that end goal.

We all soon will witness the death of this failed experiment called PFI.