Most internet companies have a wide variety of sites linking to them and sending traffic that get tagged as “referral” in Google Analytics. While this data is interesting by itself, it can become much more interesting and useful if you segment it further. Most importantly, it can help you further optimize your online media efforts. In many instances, referral sites are a combination of organic referring sites and paid referring sites (or sites that have been targeted in marketing efforts). Tracking the Return on Investment (ROI) of paid referring sites or sponsored links is a little more difficult than standard SEM because marketers with SEO knowledge know that they should keep their paid links as clean as possible without any kind of tracking parameters in order to improve search engine rankings and avoid URL confusion. This article will show you how to track the direct benefit or ROI of certain types of referring sites by using filters in Google analytics. For example, I run a photography network which has a wide variety sites linking to it. There are partner links (reciprocal links), sponsored links, social media sites, links from blogs, and other “organic links” that I cannot really put into a specific bucket. In an effort to estimate the ROI and effectiveness of my marketing efforts in these separate areas, I decided to split these referral sites into these 5 different groups as shown below (starting with #2): As you can see, by splitting the referral sites into these groups, I can see the conversion rate of each group separately and see which types of links are most worth pursuing (from a direct traffic perspective). By using this method I found that links from blogs are extremely effective at a 1.8% New Member conversion rate compared to the other types of referring sites. The 2nd most effective referral site type is social media at a 0.64% conversion rate. Also, since some of these link groups have a cost, this method allows me to multiply the visits by the New Member conversion rate, apply a value to the conversions, and arrive at the ROI for each link group. In order to do this, I set up five custom filters within my Google Analytics account which is found under Analytics Settings > Filter Manager > Edit Filter. Each of the link groups uses the same basic filter setup just with a different regular expression in the “Field A -> Extract A” field (shown below). Here is the regular expression I used to create the “Blogosphere” group of sites (you may want to add to this depending on what referral sites you consider to be blogs): blog|wordpress|typepad This regular expression tells the filter to apply the “blogosphere” medium tag to anything with “blog, wordpress, or typepad” in the campaign source. The pipe symbol, “|”, symbolizes the logical expression OR. Here is the regular expression I used to create the “Social Media” group of sites (once again you will probably want to create your own unique version based on the social media sites you find meaningful): Facebook|myspace|msplinks|flickr|twitter|photographyvoter|linkedin|delicious|youtube|propeller| fark|reddit|digg|mixx|deviantart|redbubble|smugmug|webshots For the “partner links” links group, I simply add any sites that I am currently partnered with in to the regular expression in the same manner (picturecorrect.com is an example of one of my link partners). For the “sponsored links” group, I add any sites where I have links with costs associated to them in to the regular expression. The “organic links” group is made up of any site that is left over from the various other groups (just a filter to rename the default “referral” to “organic links”). Once the filters are set up and you have data flowing in, here is how you find this report. Log into Google Analytics and navigate to Traffic Sources > All Traffic Sources > Show: Medium. When you arrive at the report shown above, you can also drill down deeper to see the sites within each link group. This method has been a huge help to me in building a successful social network and allows me to focus my marketing efforts on the most efficient sources of potential new members.