These days, you hear a lot about the importance of social signals on search engine rankings.
In the old days, search engines ranked pages entirely on the basis of on-page factors – which boiled down mostly to keyword density, and how it was applied to important elements of the page; the <Title> and <h1> tags, <META> tags, body copy, ALT text and so on. There is a fairly obvious problem with this method; it’s relatively easy to reverse-engineer the top ranking pages, duplicate the results, and obtain a top ranking position for yourself (hands up if you owned a copy of Webposition Gold back in the day).
Then along came an upstart search engine called Google. They did things entirely differently; they ranked web sites based in large part upon in-pointing links. These links were seen as a sort of endorsement from the “linker”. In other words, if Site A links to the similarly-themed Site B, then Site A is essentially providing a vote of confidence for Site B. And, since Site A contains high value and is credible content, Site B must be, by extension, credible too.
In many ways, this strategy vastly improved the quality of the search results surfers could expect to attain – at least for a time.
There were two major problems with the original Google algorithm.