Using canonical tags to help gateway domains04/21/2011
We recently purchased 22 domain names to be used as gateway Web sites for a new project. A gateway domain is used to direct traffic to a main site. In our application, the client needed to indentify traffic from different television infomercials (e.g. session starts/visits, page hits, path analysis, shopping cart abandonment, e-mail opt-in, online and offline order data, etc.). All of the gateway domains (e.g. gatewaydomain.com) redirected to one main site with a different querystring parameter (e.g. siteid=4, siteid=5...). To accomplish the job we used Response.Redirect() to move traffic from gatewaydomain.com to domain.com?siteid=_. The querystring parameter is used to track user behavior and to set a distinct 1-800 telephone number on the base Web site.
A pitfall of this technique is that Google, MSN, and Yahoo! view domain.com?siteid=4 and domian.com?siteid=3 as seperate pages. In our application both URLs serve the exact same content with the exception of a different phone number. Because the content is identical we needed to address the duplicate content issue. Making matters worse, Google indexed the base domain with the querystring (e.g. the search engine results page [SERP] or organic listing displayed as domain.com?siteid=6). Obviously, we could not have this because it would overstate traffic from an advertising source and essentially nullify the user tracking.
To resolve our issues, we used the canonical tag. A canonical tag is placed in the tag and looks like this:
<link rel=" canonical " href=" http://www.domain.com/landingpage.aspx " />
Google, Yahoo! and Microsoft have all united on the canonical tag because engines want to index and display the original or "canonical" version. Only use the tag on pages within a single site (subdomains are fine). This tag operates for the search engine in a similar way to a 301 redirect. We've been using the tag heavily since January 2010. It is now a default or standard tag that we included in all new Web sites.
When a search engine sees a canonical tag it knows to index the href specified in the canonical rather than the URL shown on the browser line. In our application, use of the canonical tag stripped the querystring parameter from the SERP and immediately fixed our tracking problems. The technique can also help keep Google, Yahoo!, etc from indexing any URLs with tracking querystrings like "?utm_source=email&utm_medium=ppc&utm_campaign=MyNewTest" or session IDs like "?sessionid=A35CA52CVAR6321VTSHG". Even if Google has already indexed http://www.mywebsite.com/landingpage.aspx?siteid=4&sessionid=A35CA52CVAR6321VTSHG&utm_source=email&utm_medium, addition of a canonical tag will rectify the SERP quickly.
Cheers to the search engines for providing us the canonical tag.