An unofficial blog that watches Google's attempts to move your operating system online since 2005. Not affiliated with Google.

Send your tips to gostips@gmail.com.

March 11, 2008

Google Sitelinks Using Traffic Stats?

Google's automatically-generated sitelinks are not always useful and sometimes they look very weird. According to Google, sitelinks are displayed below some sites in the search results and they "are meant to help users navigate your site. Our systems analyze the link structure of your site to find shortcuts that will save users time and allow them to quickly find the information they're looking for."


The first sitelink for blogger.com uses the following anchor text: "Thomas déjeune 1" and sends you to this broken link: http://www.blogger.com/post-create.g?blogID=XXXXXX. The second sitelink uses "Diseño" ("design" in Spanish) to send you to another broken link: http://beta.blogger.com/rearrange, used by the most recent version of Blogger to access the layout editing page. These two links have a lot of things in common: they use strange anchor texts, they link to pages only accessible when you're logged in and anonymize the personal information (the blog ID).


The first sitelink for Google Toolbar is a link to a page that shows up in a single context: after you install Google Toolbar 4 for Internet Explorer.


Most of orkut's sitelinks are pages inaccessible without authentication and the most popular feature of Google's social network is the scrapbook.

These examples could indicate that Google started to use traffic statistics, probably gathered from Google Analytics, Google Toolbar users or from third-parties. In a patent that describes the sitelinks, Google mentions the toolbar as one way to acquire traffic data.
In one implementation, assume that users have downloaded/installed a toolbar on their respective clients that facilitates web searches on a search engine, such as search engine system. In this case, the toolbar may include software code that instructs a client to send hypertext transfer protocol (HTTP) requests to server for each web page that client accesses. [The front end] may use the information in the HTTP request to identify the particular web page and web site associated with the web page that client has accessed. Alternatively, [the front end] may receive similar data when clients click on links provided by search engine system.

5 comments:

  1. Google also is collecting traffic stats from those who have installed Web Accelerator.

    ReplyDelete
  2. Interesting observation. Consider this search which returns one of my company's websites at the top:

    http://www.google.com/search?q=free+translation

    The first two links are for pages which only work when you do a POST or GET. People often link to them after receiving their free translation thinking that it will be a permanent link. I would've thought it's getting these links from other websites linking to them. (And if it's getting this data from Google Analytics or Google Toolbar, I wonder where it's getting the anchor text from.)

    ReplyDelete
  3. Just want to ask, Google create sitelinks for Blogger blogs as well or just high PR domain names only.

    Stay Updated With Technology
    http://feedshub.blogspot.com

    ReplyDelete
  4. Thanks for the posts about sitelinks. I have been trying to get resource about google sitelinks. I have also experienced a regional search difference for the sitelinks results.
    I did try the google search for "javra" aswell.

    --
    abeen
    http://abeen0.blogspot.com

    ReplyDelete
  5. I am pretty sure Google isn't looking at traffic metrics to determine sitelinks. From everything I have read they are working on a method to allow the webmaster the means of defining it through the webmaster account. But, right now all you have to do is to architect the URLs using a href with alt text and block sitelinke you don't want to show up.

    ReplyDelete

Note: Only a member of this blog may post a comment.