In Yandex.
In google.
Now, how to read these graphs.
Was taken by total weight of links bought for 18 months. Total - slightly more than 18,000 references. This is so that there was no reason to doubt the representativeness of the results. Further, all references were checked for the index in Yandex and Google. Please note that all links were bought on the basis that the page where they are located - in the index of search engines.
Then link each month were divided into two piles: those that are already in the index. And those that are not yet in the index. Not a page that is already reference. If anyone knows, the indexation is conditional links, if it gets into the page cache, where it is. T. e. also in the index page. Schedule a handful, which is the index we can see in the picture.
Yandex:. approximately 7 months, the percentage of indexed links about the same, then begins a sharp rise, and the longer the reference sticks in place, the more likely it will be in the cache Yandex. In simple terms, Yandex tupit and assumes no references in the index until they are at least 7-8 months. He then refers to those links more trusting. You can make more profound conclusions that the longer the reference is, the higher the trust, which means that Yandex starts in the bulk to consider only those links that have upheld a year and a half, and the first 7-8 months generally considered only one-third the mass of the reference.
Google:. the picture is quite different. First, Google quickly indexes the links, and then begins to throw them away from the index, along with pages. A year later, began to return to the index, reaching the initial value of 40%. If you still build a profound conclusions, then Google would be like trying to play faster referential fresh weight, and then, later, the same 7-8 months, it starts to sort and discard excess. What begins in a year can be characterized as a possible set of confidence on the part of Google's.
For some reason, there is confidence that the rental references ( sapossylki ) have similar distribution.
How to use this to work? . To speed up their work has meaning, still forced to speed up the indexing.
In google forced acceleration does not make much sense. Links that do not change their place will be more than a year in Yandex even more than in Google. For Google's more important, fresh weight, rather than for Yandex. The last conclusion is controversial - it is rather the projection of the findings from the study because it remains unknown and the weight of links, and how it grows with age. But if you find a better reference indexes, standing for longer, then in theory it should, and in other ways, to love them more. As well as the reverse argument for google.
More information on the conference. All-in-Top 2012. February 1-2 in Moscow. And I continue to turn the numbers. Some figures are quite interesting.


No comments:
Post a Comment