Here, we examine how Insight Whale increased the conversion rate of an online pharmacy based in the UK. The focus was on the click-through page for erectile dysfunction rather than ecommerce conversions, due to a long sales funnel. In the case study, we examine the use of a two-phase experiment and how it facilitates working as a team with our clients.
The main objective was increasing click-throughs to the next page and as a result increase of overall conversions into sales. Ecommerce conversion was not the main objective because the purchase funnel in this case was a long one (click-through-page → product page → medical conditions questionnaire → checkout page) due to peculiarities of selling medicines online. As the client was using AdWords to drive traffic, he had to create click-through page (which didn’t include names of the actual medicine being sold) to land traffic on.
In this case I used pretty much the same ideas I used when optimizing online apparel store product page: using decision making funnel and adding trust-signs, crucial for conversion.
There were 2 phases of the experiment.
In the first phase I wireframed new variation and my developer implemented in Optimizely. This experiment lasted almost 3 weeks and showed pretty good results: 26% growth of click-throughs and 12,6% ecommerce conversion rate growth.
As not all experiments turn out to be successful, I always follow lean approach in A/B testing. One of the characteristics of lean approach in A/B testing is not using a designer while creating variation, but just making a good draft, which may not look pixel perfect, but works good for the sake of experiment.
So, when we got the initial results, the client decided to engage a designer to prettify our variation and see if it will help increase conversion more. And it did. Though the client was a bit impatient and stopped the second phase of the experiment after 5 days, the prettified variation showed good results over the draft one.
The 2-phase structure of the experiment is a good example how important is to work in team with the client.
One more thing to notice in the experiment is how my developer implemented the variation in Optimizely. The page was fully reorganized using jQuery (the actual development was made outside of Optimizely interface and only after the variation jQuery code was created, we added it to the Optimizely code editor). You can ready about our approach while building A/B testing experiments for Optimizely and Visual Website Optimizer with jQuery here.