Wednesday, June 18, 2008

Google Analytics - not really working, yet

For a little over a month, we've been trying to implement google analytics on Time4Learning. So far, I don't really feel that we've learned that much. The good news is that:
- the number of conversions from Google Analytics matches (within a few percentage points) what our shopping cart/credit card system is showing
- we are beginning to get a sense of which of our referring sites (aka ads and partner sites) are contributing not just in terms of traffic but conversions.

It just occurred to me that we don't know if the we are correctly tracking the PPC traffic. We installed tracking codes on our Google, Yahoo, & MSN PPC campaigns so we can track that traffic separately. I wonder if the the PPC results from within those programs matches the Webanalytics results?

The big problem is that 70% of our conversions are being tracked by Google Analytics as "direct", meaning they found our site through typing in our URL or through a bookmark. I simply don't believe that since our users report, as part of the signup process, how they heard about us and we get around 70% reporting that they found us through the search engines.

I also noticed that we got NO conversions out of our top 16 keyphrases from natural search.
This feels unlikely to be true.

When we dug into the statistics, I was told because of the iframe coding on our demos, the sessions could not be tracked. (which is why the iframe demos were recoded with java script).
This would explain why two thirds of our conversions are classed as "direct entry".

But, this doesn't make sense. Google analytics, much like PPC, is primarily cookie-based. This means that whether the session was trackable or not, google analytics ought to be able to track the conversions back to the original entry to the site. So even if a user found us through natural search, bookmarked us, return a week later and signed up, Google analytics should report them as natural search.


No comments: