It's The End Of Google Website Optimizer -- And I Feel Fine
If you’re a Google Website Optimizer user like me, you likely received an email last week indicating that the free Google testing tool was being discontinued. While at first I was disappointed by this announcement, reading further, Google said it would be integrating testing functionality directly into Google Analytics.
For those of us who have used Google Website Optimizer for many years, the change may seem challenging. However, upon further review of the new Google Analytics integration (deemed “Experiments” and found under the Content section), there are two main benefits I already see from the new system.
1. Encourages goal creation in analytics. I work with a variety of clients, and very few of them have set up goals in Google Analytics. I always feel surprised by this, because goals can be very simple to set up and provide marketers with such rich data about tracking conversion actions on a website.
The new Experiments testing tool relies on goals as the target conversion to test success or failure of a given set of pages. There’s a very easy dropdown menu to select the goal you want to target for that test. Whether that goal is newsletter signups or ecommerce sales, you can track each individually. Hopefully, this will help encourage marketers to employ the use of goals more often in Google Analytics, for use with Experiments as well as for the other benefits of receiving goal data.
2. Track different types of goals. One limitation of the old Google Website Optimizer tool was that you couldn’t track events as conversion actions. Instead, you were limited to tracking an entirely separate URL as the goal page. No longer! With Experiments, you can now track event goals, which might be used on an embedded video. For example, if you want to track how many people click on the play button to watch a video, you can now track that as the goal for the page test. Pretty neat.
What’s Still Missing…
At first blush, it doesn’t appear that ecommerce transaction values can be tied to Experiments… yet. One of the major challenges I had previously with Google Website Optimizer was that, for ecommerce clients, the quantity of conversions wasn’t always the main success factor. Rather a more accurate success factor for most ecommerce sites is revenue. The old tool focused solely on quantity of conversions, so an ecommerce site owner couldn’t readily tell which A|B page version drove the most sales revenue. However, using the Content section in Google Analytics and comparing the two pages could yield the result. But if ecommerce sites want to truly test revenue generation from page variations, Experiments will need an update.
Multivariate Testing Removed
The downside of the new integration appears to be that multivariate testing options have been removed. However, you can still perform multivariate tests; you just have to create separate, whole pages for each variation, which can be much more time-consuming than the old Google Website Optimizer method of adding variations via code on just one page, allowing Website Optimizer to switch out variations for you.
I wouldn’t necessarily worry too much about the loss of true multivariate testing, though. In many cases, I’ve found that clients rarely have a sample size large enough to test all of the variations in a multivariate test fully. For me, A|B has always been the preferred testing method anyway, because I can often get fasters answer with less variation combinations to test at once.
I’m looking forward to trying out the new testing tools in Google Analytics. If you have some thoughts on it, I’d love to hear about your experiences. To learn more about Experiments in Google Analytics, visit the help pages for Google Analytics.