This will only be a shorter post than normal because it doesn’t apply to everyone – well mainly to those who use google analytics with sites where their pages have unintelligible endings looking something like this www.yoursite.com/default.asp?docId=15097 (usually running on cmses) or running on different subdomains or domains eg www.yoursite.com/analytics or blog.yoursite.com/analytics who want to have clean logical webpage names in their reports.

In this post I will describe three things:
1. How to use google analytics advanced filter so you can see the whole uri including host name in your reports
2. How to add urchin tracker to your web page code to give your pages logical web page names
3. How to name your webpages carefully

1. If your site has different domains (I’ll do another post on this later but if your site is running across different domains you’ll need to use the 3rd party shopping cart and utm linker or as visitors go across domains they’ll be treated as a new visitor each time so inflating your numbers) or subdomains, you won’t be able to see the domains in your GA reports. The reason being by default Google analytics only picks up the last bit on your uri so you can’t drilldown and see which pages come from which domain or subdomain. In addition, if you have the same uri ending eg /analytics, GA will be lumping them together (even if the reality is these are directing you to different pages on the internet). 

For sites with different sub domains and or domains, I use a Google Analytics advanced filter to allow me to see the full host/domain name. If you try to work out how to do it from GA’s documentation – you might have a bit of a hard time here so I have spelled it out here – see Robbin Steif’s quite fab GA “worst of” documentation competition if you don’t believe me. And on a similar note, I am really looking forward to Justin Cutroni’s new pdf book called Google Analytics Shortcuts which will cover every imaginable detail on link tagging, setting up shopping carts etc.

First in GA click on the profile settings of your site and then edit filter.
Under filter type choose – custom filter from the drop down menu and tick on advanced.
Field A > Extract A – Host name – (.*)
Field B > Extract B – Request URI – (.*)
Output to > constructor – Request type URI – $A1$B1
Field A required – Yes
Field B required – No
Override Output Field – Yes
Case sensitive – No

 Click here to see the advanced filter area in GA

What this does is extracts all of your host names and outputs them to all uris requested (.*) (all regular expressions) to all of the fields ($A1$B1) so that you end up being able to see the uri including domain in your reports.

2. It’s really important for your web pages to have logical and understandable page names on google analytics reports so that you can easily see what pages people are going to on your site ie which is more helpful books/analyticsbooks/GoogleAnalyticsshortcuts or /default.asp?docid927196? If your site runs on a cms or for whatever reason does not have logical page names then you can add the urchin tracker javascript tag at the bottom of your webpage code to call the page name you include in the tag. Increasingly as url rewrite engines are used to give logical page names to sites, this will not be necessary – and definitely consider a url rewrite engine as a neater solution . On the other hand, setting up a GA advanced filter to replace the default URI with page titles is no help as page titles will have been search engine optimised. But assuming your pages are full of /docid74363494 then use the javascript tag to give your pages the names you want in your reports.

Here is the javascript urchintracker tag where the area at the end in brackets is what you decide to name this particular webpage e.g books / analytics book / GoogleAnalyticsShortcuts. Add the tag to the bottom of your webpage code just before the body tag. The first name in brackets being the section of the site, the second name being the site subsection or product details and the third the current webpage – but these can be defined by you:
e.g script language=”javascript”>urchinTracker(”//books/analyticsbooks/googleanalyticsshortcuts”)

3. You need to give your pages logical, meaningful but concise names that reflect their position in the site. The easiest place to start is to look at the wireframes/site tree and define page names according to the section of the site (eg books – quite general), then analytics books (the type of books or products) and finally Google Analytics Shortcut (the actual book name or product). So we end up with, (”//books/analyticsbooks/googleanalyticsshortcuts”). The final name is the current position on the site. Do not use spaces or other special characters such as ‘&’, ‘%’ or ‘+’ in your page names or the tag will not work. Whilst setting up a list of pagenames, map out all of the site pages with the help of the site tree/wireframes to avoid using page names that been used elsewhere on the web site. Also consider prefixing for example your FAQ page with a unique identifier – say, ‘analyticsfaq’.

To conclude

GA advanced filters can be fab once you know how to use them and their documentation is famously lacking (no offence taken I hope). In this instance being able to see host names (domain) means you know which domain particular pages come from as well being able to drill down into specific domains. The javascript urchin tracker tag, although not the neatest of solutions, and for large sites definitely investigate a rewriting engine pronto, gives you the wonderful ability to give each page the beautiful logical name it always wanted to have. And for the purposes of GA reports, having all of your pages beautifully named across top content reports and in which ever way you choose to segment them, has to be a good thing.

I welcome your feedback, agreement and disagreement. And if anyone knows a better way of doing of these things then please do let me know.