Campaign Tracking Beyond Landing Page with Google Analytics. Part Two – Segmentation
To quickly recap part one of this two part post… Google Analytics is great at showing how many visitors, page views and what content web visitors are looking at but, falls flat when trying to track paid campaigns. Paid campaigns include links from e-newsletters, paid links from other websites, and of course banner ads. Part one describes how to use the Google URL builder to tag links. (Note if you’re using Google AdWords there is no need to do this since it’s done automatically for you.)
Once the campaign has started (and someone clicks on a link) results will begin to show up under the Traffic Source->Campaigns section of Google Analytics. If you have more than one campaign running this will tell you which one is bringing the most clicks to your site.
In the example from part one I created a fictional campaign called marketing2009. The campaign has now been running for two months and results so far show that the link was clicked on 11 times. February 11 was the day with the most clicks and then after that no more clicks. Great… if that’s all you’re really looking for. But what did these people do once they came to the site? Did they buy a left handed widget like you wanted them too? Did they sign up to receive a newsletter? Did they become a lead or did they bounce off the site? In other words did they convert? This is where segmentation comes in.
Google introduced segmentation to Analytics at the end of 2008. Prior to the introduction there was no really easy way to slice & dice your traffic. Now cutting up data is a breeze. On the left hand nav bar in the “settings” section click on Advanced Segmentation. Using the Marketing 2009 example, I’ll created a segment which will look at the 11 clicks we got from this campaign. The first step is pick which data point / dimension or metric to segment traffic on. In this case it’s the Marketing2009 campaign which is found under Traffic Sources. Drag the green campaign box to the dotted outlined box.
Next in the drop down box labeled condition pick Matches Exactly. Next enter in the name of the campaign we want to focus on. You can type a few letters and Google will pull up campaigns as you type. If you want to add another data point to segment on you can do so by clicking the “Add Or Statement” or “Add and Statement”. I want to keep this example simple so I’ll give this segment a name and click the create segment button. I’ve not hit an upper limits of the amount of different metrics you can have in a segment, however, the more you add the less & less data will match until you have no data in your segment. Click the Test Segment button to see how many visits fall into your segment.
Once you’ve saved your new segmentation you will see a list of all the segments. Default segments which Google creates for every session are at the top. User created segments are at the bottom. Find the segment and click you wish to view and click the APPLY TO REPORT link. Google now opens the dashboard page as if you’ve just logged in. However, the numbers shown are only for our Marketing2009 campaign. Now I can see that of the 11 clicks that came from the campaign there were 151 page views, that each visitor stayed on the site about 30 minutes and the bounce rate was a low 18%. Clicking onto traffic sources again I can see which ad brought the most people in. (if there were more than one ad). I can see what the most popular pages on the site were for these 11 people, and how long each page kept their interest.
Again this is all fine and dandy but the reason you’re running a paid campaign is to get traffic to your site so that they will do something that you want. In this example I’ve setup a form which I want people to fill out. Of the 11 people who came in only 1 filled out the form. Therefore, only 1 person converted and my conversion rate is 9%.
Paired together, segmentation & URL tagging can get your past basic reporting of basic page views, visits & clicks of how well your paid campaigns are doing.
Leave a Comment
You must be logged in to post a comment.