Following the latest posts on time on site on this blog and on Occams Razon by Avinash Kaushik, I received a few questions and I also saw a few comments about Segmenting Visitors by Average Time on Site.
On my last post I demonstrated the results of such a measure – on this post I will write the exact cook book which enables this task.
Step #1: Get ClickTracks Analytics
ClickTracks is a powerful tool which main edge is its ability to segment users by their behavior patterns. ClickTracks will work either with log files or with a javascript and all you need to do is decide what’s best for you, then to download a trial on their web site: http://www.clicktracks.com
Step 2: Configure your Dataset
Build a profile with clicktracks. As I don’t know if you chose logs or tags, I will not get too deep into that but guide you to the ClickTracks Way help Files instead.
Step 3: Review the “Site Overview” Report
The “Site Overview” Report enables us to see the main analytics KPI’s such as
visitos, pages, keywords, referrals, geographic parameters etc. One of the most imoprtant KPI’s of this report is “time on site” and “page views per visit”.
Step 4: Create Segments!!!
Using “tools” > “Advanced Segments” > “Session Length” > “At Most” 5 Seconds > 4 times “Next” > “Finish”
We have now created a segment for those visits where the user spent at most 5 seconds n the web site.
Next thing, is to build another segment usin the same method, only this time we will select “At Least” 5 Seconds.
We can now see the different in behavior between 3 groups:
1. All Visitors
2. Short Visits – Those who don’t really see our web ite (what can you see in 5 seconds?)
3. Long Visits – Those visitors who really see the site. this number is the most important one and actually means how much time our REAL users are spending.
We can later on compare referral sources, Keywords, Search Engines, Geo-Targeting, Campaigns and to get insights on the various characteristics of the traffic we bring in – therefore to come up with practical ideas for optimization!
One thought on “Segmenting Visitors by Average Time on Site”