Google Website Optimizer: A Eulogy…

If you’re a user of Google’s excellent Website Optimizer (GWSO) tool, chances are you’ve already received a message notifying you that as of August 1st, our beloved GWSO shall be no more. This particular Tech Guy is taking it rather hard. That’s right, GWSO is being retired and…changed.

Some might say upgraded. Time will tell. Missed it shall be!

There. Eulogized. Sniff sniff. Time to move on…

Obviously this is going to cause some headaches and major changes to the way that we go about testing within Google. All testing is going to be done directly WITHIN Analytics, which is a good thing. Unless you hate the newer layout, which hides things a couple extra layers deep…then you’ll hate it. Both are quite acceptable feelings.

Optimizer tests are now called “Experiments”. They can be found by going to Content -> Experiments. You’ll be prompted to setup your first experiment to get it going.

The setup is actually very similar to the GWSO setup, but it’s been changed. There’s no more multivariate testing. It’s straight up A/B testing only now. You can choose up to 5 test variations to run against your control page. Experiments use Analytics Goals to track conversions, which some will find is a nice change of pace. You can use either URL Destination or Event Goals to track conversions. Adding Event Goals is useful for beginners that didn’t know that this was actually already built into GWSO, you just had to ignore their warnings, thanks page, etc. and march to your own drummer. (Note: You CANNOT use Average Visitor Duration OR Pages per visit in the current version of Experiments.)

So what does this mean for you? It means it’s high time that you set Goals and started prepping for this important change! Decide what you want to track and what you want to test. Now that it’s all integrated under Analytics you don’t have much of an excuse for not setting and tracking your goals.

It also means that you’re likely going to need to learn exactly how to properly setup the correct tracking scripts for things like Cross-Domain tracking (use a third-party shopping cart?) and ensuring that your analytics sessions don’t mysteriously get butchered by some rogue page programming or redirection, (it happens) or by jumping domains and ending your sessions. For most testing that shouldn’t be a problem, but there’s lots to keep in mind when working with Analytics and its infinite funness!

If you’d like more information or need help making the transition, let us know! Our robots are standing by.

You can find the complete walk-through for Experiments here.