I started this with the original post cleverly named Campaign Stacking in Omniture but wanted to expand a little based on some emails I got from people wanting to hear my experience with it. Since the post in December I implemented the campaign stacking plugin but haven’t really exploited the data that was captured. Couple of reasons why:
1) There still isn’t a great way within SiteCatalyst or Discover to visualize the data as to how visitors are encountering tactics, the typical patterns, or the influence of the tactics on one another. Basically we are just capturing the string and then classifying the tracking codes into categories. The usability of the data is still needed. I wanted to bring that up at the Omniture Summit in March, but we ran out of time, and I’m sort of shy.
2) Classifying the data is time-consuming. Since every string of unique tracking codes is captured you get a ridiculous amount of data to classify. This takes forever even after I built some databases to automate some of it.
3) Doesn’t capture things that aren’t campaign identifiers. For example, we use the Genesis integration for DoubleClick which passes data from DART but not as a cid. The campaign stacking logic is looking for a cid= in the url string. DART doesn’t have a cid. Also, you miss out on non-campaign tactics like SEO or people just coming in via Bookmark/Typed in.
4) The last limitation is that you have no frame of reference as far as time goes to the campaign stacking. You don’t know the time between each tactic. Like is it 3 weeks later that someone goes from Paid Search to an Affiliate? There is no timestamp captured.
If Omniture fixes all of these things, Campaign Stacking becomes infinitely useful.