How Drukwerkdeal Used A/B Testing To Get An Increase In Purchases

The main objective of this optimization campaign was to increase sales by changing cross-selling elements on product pages.

Paul reviewed a number of product pages on the original website and realized that the cross-selling messages were not convincing. He feared that this was doing them more harm than good.

This is how one of their product pages looked:

Notice the links in green under the product description that were intended to cross-sell.

Paul decided to remove the cross-selling links and run an A/B test against the original to see what effect this change would have on their sales. Since this was going to be a big move, he decided to first test the change on only a few product categories. To implement this, he used pattern matching to include some URLs on which they wanted to test.

The variation product page after removing the cross-selling links looked like this:

More than 14,000 visitors became a part of this test that ran for 2 weeks. The metrics on which this test was judged were average order value (AOV), number of products purchased per order, add-to-cart conversions, and transactions.

All test data was pushed into Google Analytics (GA) with a single click while setting up the test on VWO. This allowed Paul to obtain holistic insights on a range of parameters apart from absolute conversions.

The variation page recorded 5.6% improvement in the number of online orders completed. The test data also helped Paul to conclude that the variation page performed better for both the new and returning visitors and all traffic sources, particularly for organic and direct.

To quote Paul, “We do a reasonable amount of testing on our website and try to be very curious about all the things we add to our site. I had a sense that it would distract visitors and would have a negative impact on conversions. Now we know that it did.”

Bolstered by this success, Drukwerkdeal implemented the variation on all its product pages.