When has google taken account of page changes in the Serps.
Is it when it crawls the new content?
Is it when the new content appears as googles cached page?
Or is it some other time?
tedster
4:40 pm on Sep 4, 2010 (gmt 0)
As I understand it, we can think about four different actions.
1. Crawl the URL, storing the raw data in Google's back end shared cache (private)
2. Shard the raw crawl data, computing the ranking factors in the new content and storing those results as meta data assigned to the main URL record
3. Integrate the private meta data into higher level master files used for ranking calculations
4. Update the public cache
The new Caffeine infrastructure had a powerful effect on #3 - which is where visible ranking changes happen. With the old Big Daddy infrastructure, crunching new meta data had to happen in big batches across a major chunk of the index at one time - in other words it waited. With the caffeine, the new ranking calculations can happen in a much more granular fashion.
Some ranking calculations have time delays and trust checks associated with them, so #3 does not always integrate the full effect of newly found changes on the first calculation.
The public cache dates (#4) do not have a direct relationship to #2 or #3.
Mark_A
8:28 am on Sep 9, 2010 (gmt 0)
Hi tedster, that is interesting.
I have recently done some sitewide changes like changes to all the H1 across the whole site (plus changes to many page titles) and adding footer links across the site etc ... at the moment I can see that in some cases the google cache has the new page, but also in many cases the rankings have not changed at all, they are still bang on where they were before.
