It took no real science to convince us that Responsive websites would outperform their non-mobile-optimized brethren. We’d made that leap of faith (more of a small step, really) long ago. Without objective data, though, it’s hard to separate the facts from the closely held beliefs.
We should start by noting that optimizing sites for mobile screens takes time. In order for a site to look and perform well on a wide range of screens and devices, a fair amount of work needs to go in to designing the site to, well, respond to those conditions. That time adds cost.
So, with little data supporting our belief and more cost required to perform the work, we had set ourselves up for a recurring conversation with clients. “We know this will help you, you’ll just have to trust us with a bigger part of your budget.” You could say that just about everything we do has to be taken on some measure of faith. A lot of our work happens before a site ever launches. But, it’s always easy for a client to view a new technique or model as an add-on. Especially in the case of ecommerce, where the site’s explicit purpose is to drive a return on investment.
So, we ran a test.
We took a popular ecommerce store (O’Neill Clothing) that we’d recently redesigned and monitored conversions, transactions and revenue for three weeks. Then we quietly deployed the responsive conditions to the already live site and monitored for another three weeks.
This was not an A/B test. We simply picked 6 non-holiday weeks that perform similarly year over year to get as near to similar conditions as we could.
The “responsive conditions” were typical mobile patterns. We made the site fluid. We collapsed the primary navigation menu, allowing visitors to expand it by tapping a Menu link. We increased the size of the font, the tap areas and detail photos. We reduced the number of columns. We spent a lot of time just “fixing Magento forms.” Everything in a way that lets the O’Neill team continue to manage 100% of the content on the site.
Here’s what we found:
CONVERSIONS: + 65.71%
TRANSACTIONS: + 112.50%
REVENUE: + 101.25%
CONVERSIONS: + 407.32%
TRANSACTIONS: + 333.33%
REVENUE: + 591.42%
If you don’t spend a lot of your own time swimming in ecommerce analytics, transactions is the most important number to watch. iPhone transactions more than doubled. Android transactions more than quadrupled. All from relatively quick optimization.
Our work here isn’t the best example of Responsive Web Design – this was a case of Mobile Second as the secondary navigation will clearly demonstrate – but that’s part of what we like about the model. It isn’t all or nothing. Because there aren’t multiple versions of the site, we can make small tweaks to an existing site without much effort. In a matter of a few days, we were able to take a large, active site built on a famously stubborn ecommerce platform and dramatically increase mobile usage. And the best part is that “mobile usage” refers to real people, people looking to buy boardshorts and backpacks.
How you like them apples?
UPDATE: we’ve posted some additional information on non-mobile traffic during the test period.