This article was published on Apr 27th, 2012
As a follow-up to my series of articles on the latest Google Panda Dance, I thought it would be a good idea to show a video made by Matt Cutts, (Google Search Engine Marketing team) in which he answers several questions about the public about:
How Google Ranks Content.
The video is 8 minutes long and in it, Matt goes into some detail about various aspects of the methodology of Google Search.
In this video, Matt answers a question related to the process of ranking and website evaluation, starting from the crawling stage, to the analysis of a site, to crawling frequencies…
Google main priorities in their drive to be the best search engine they be are based on 3 principal factors:
- Crawl the web for existing and new pages
- Index these pages
- Rank these pages so that the most relevant are served first.
1. The crawling process
Starting at 1:15 min
In the “old” days (the year 2000), it took google 30 days to crawl, one week
to index and propagate to their data centers (another week) and this process
was called the google dance! In 2003, Google switched to a different format
and began to crawl on daily basis which is substantial departure from their
original method lasting 30 days. This allowed them to refresh their data banks
on a daily basis, and thus be kept more up to date to changes. Over time, they
have improved their incremental updating of their index.
2. The indexing process Starting at 3:55 minute. Here is where
it becomes interesting as it gives a very good insight into the google indexing
mindset. Given that indexing is taking things in word order, here is how it
works. To illustrate this process, Matt Cutts give the following two words as
an example of a Google search.
Katy Perry
In the crawl, Google established in which document each word was found. In
other words, in which document does the word Katy appear, and in which document
does the word Perry appear. So for example:
| Katy appears in the following pages | Perry appears in the following pages |
|
1 |
2 |
|
2 |
8 |
|
89 |
73 |
|
555 |
555 |
|
789 |
1000 |
So when someone types in Katy Perry in Google, the search engine sets upon
finding out in which document do these words appear, and it does it like this:
Document 1 has Katy and document 2 doesn’t so doc 1 is out.
Document 2 has both Katy and Perry, so that document is one of the possibilities.
Document 89 has Katy and document 73 doesn’t so doc 89 is out.
In other words, only document 2 and document 555 have both words.
So Google searches for pages and back links (anchor text pointing to that page)
where both words appear. When this document selection is accomplished,
not the ask is to figure out a way to rank them.
3. The ranking process
Starting at 5:45 minutes
This process is tricky and Google uses Page Rank (PR) in combination of some
200 different factors to determine which page to rank highest.
If this video has shown anything, it is that
- Page Rank is a determing factor as to how well a page can be ranked
- Page Rank is dependent upon the reputation of a site, the number of back links to that site, and the reputation of those sites from which the backlinks orginate!