Marketers generally aren’t at a loss for ideas. They have thoughts about messaging, new graphic treatments, pithy ad copy, site redesigns, brand refreshes, you name it. But sometimes, they get stuck worrying about how to fix something instead of wondering whether it’s worth having there in the first place.
Hi Chris! Can you start by telling us a little about yourself and Disruptive Advertising?
Sure, man. I’m a big, big believer in a world where every business tests their ideas. I started my conversion pptimization agency Dayley Conversion in 2014, which later merged with Disruptive Advertising. That’s where I currently work as VP of site testing and optimization.
As for Disruptive, our leadership team spent years at Omniture and Adobe. They took those enterprise best practices and built an agency that effectively applies analytics to get the most out of PPC budgets and site conversions. We’ve developed industry leading tools, systems and processes but welcome the unique challenge each of our clients bring for us to solve.
So how long have you been doing testing personally?
About 5 years now.
We’d like to get specific. Let’s talk about the some of the tests you’ve ran recently. You mentioned Musicarts.com when we spoke a while back. How did you settle on what you were going to try there?
It was a part of our existing testing strategy that we execute for enterprise level clients. In essence, one thing we’re trying to learn is what works on a page and what’s keeping a user from finding what they really want? So we systematically test removing most elements on a site to gauge the impact of the element, which we then use to prioritize future test opportunities.
Can you give us an example?
Sure. We thought removing the “Recommended for You” section might help users feel less overwhelmed, and focus attention on more relevant offers. We felt strongly about this idea since the recommendations were not very personalized anyways. Here’s the control and the test:
And how did that work out?
We saw about a 31% increase in revenue per user.
The fact that removing something from the homepage increased conversion rates told us that there was clearly too many design elements for the users to pay attention to. Further, the actual recommendations themselves were not personalized at all, and were not for the most popular or the most profitable products. The client was surprised because they figured more was better when it comes to ecommerce, and of course thought that the recommendations were for products people would want to buy, especially if they were being told they were “recommended.”
Just goes to show that you can’t assume something on a site is helping your conversion rates until you have tested the impact of having it there!
So there you have it: if your site is using “personal recommendations” that aren’t actually all that personal, consider removing them. Users might just find what they’re looking for a little faster.
In the example above, Disruptive got massive lift making just one simple change. With Sentient Ascend, you have the ability to test dozens of ideas in thousands of potential combinations all at the same time. It’s a whole new approach to CROs and it can get you to better results faster than ever before. We’d love to show you how it works. Get in touch!