A big question in the Magento community is, How do I improve revenue and conversion rate on my Magento eCommerce site? The best answer is conversion optimization. Conversion optimization takes your current site and tests different scenarios to determine what changes may best improve your site. Conversion testing can include making changes to layout, image, call to action or any number of other items and then testing the changes against the original to see which performs better.
This sounds great, but how does one create and manage these tests on your site? It sounds like it could get complex. Thankfully, there are several tools available to help you with this process.
While most tools can be costly or have limited free service, Google Website Optimizer (GWO) is a completely free and robust conversion testing tool. GWO provides the option to perform A/B split testing or multivariate (MVT) testing, both with native Magento integration. What more could a Magento store administrator ask for? Unfortunately, there are some limits to GWO and its native functionality with Magento, but even with these limitations, using GWO for Magento store conversion testing will provide you with many testing options and statistically valid results.
So now that you are undoubtedly interested using GWO to test your Magento store, you are probably wondering, What is GWO’s native Magento functionality, and what will that functionality allow me to test? There are three different testing categories included in the GWO-Magento native integration. Specifically, A/B split testing for a CMS page, multivariate testing for category pages, and multivariate testing for product pages. The remainder of this article will be devoted to explaining each test type. For further information, I would suggest viewing a great webinar posted on What Is Ecommerce.
A/B split testing a CMS page
Magento and GWO easily integrate to split test pages in the Magento Admin Panel. Using this function will allow you to test separate pages against each other. Generally speaking, this means you have an original page and redesigned variant test page. Using GWO, a visitor to your site will be directed to only one of those pages. Number of visits and conversions will be recorded in GWO and used to calculate the winning page.
Setting up the test is as easy as navigating to CMS pages in your Magento store backend – then entering test information in the selected page.
The best time to use this test is when you want to introduce a new element or design to a page. Essentially, this testing allows site visitors to decide if the new changes are better than the previous design.
In addition to deciding to test, it is also extremely important to choose the best conversion success page for testing. We realize that it is tempting to use the checkout page as the conversion success page for all tests, but please don’t fall into this trap. Choosing the conversion success page is very dependant upon what you are testing. For example, if you are testing the call to action or page placement of a newsletter sign-up, you would want to use sign-up completion as the conversion success page. You may also want to track newsletter site visitors to see if they convert better than regular site visitors by independently following the test using Google Analytics (or any other data collection system).
For more information on A/B split testing, you may want to check out this A/B testing resource from Smashing Magazine.
Multivariate testing a category page
Multivariate testing category pages with Magento’s native GWO integration limits you to testing only one banner and descriptive text per category. Setting up the test is very similar to setting up a CMS page split test. It requires navigating to Catalog, then Manage Categories. The Category View Optimization tab can be found when a category is selected, and test-specific information is entered here.
You want to use this test anytime you are considering making banner or text changes. Actually testing category banners can be extremely useful. For example, we collected click data for a site and noticed a number of visitors had tried to click a banner that did not have a hyperlink. Based upon this data, new hyperlinked banners were tested, which improved the add to carts for this particular site.
But why did we use add to cart as our conversion success instead of checkout success? Because add to cart was the goal metric we found useful for this test. If you think of your eCommerce site as a shopping funnel, category pages are close to add to carts in the funnel, but far from a checkout success page.
Multivariate testing a product page
Using Magento’s native GWO integration will also allow you to perform multivariate testing for product pages. You can test up to 8 attributes per product. Setting up the test is quite easy and similar to setting up a category page multivariate test, except for a few small adjustments. For product pages, instead of navigating to Manage Categories, you navigate to Manage Products, select a product, select Product View Optimization, and then enter your test information.
This test is extremely useful when changing product attributes. For example, you can find out if changing the description will increase product add to carts. *Notice I again suggested using add to carts as conversion success instead of checkout success.
While all this testing information is great, you may be interested in the average order value (AOV) for one test set vs. another, or you may be interested in add to carts vs. visitors that actually purchase. To get those results, you can use Google Analytics’ Custom Variables to track which variation a visitor sees and eCommerce data associated with that test variation.
Looking beyond testing with GWO
There’s no question that Magento’s native integration with GWO – in and of itself – can be extremely helpful with improving your store’s conversion rates, but for the best results, you might consider going beyond testing with just this native integration. Detailed conversion testing on a Magento site can be quite complicated, but more often than not, it proves to be extremely useful for eCommerce sites. To ensure testing is conducted correctly, you’ll want professional help. To get the best conversion numbers possible, I suggest hiring a professional agency to design, set up, and manage more extensive testing for your eCommerce site.
With Google Content Experiments now live, please stay tuned to the Blue Acorn eCommerce Blog for further updates on conversion optimization.