A few months ago, in one of our high traffic verticals, we decided it was time to start A/B testing. In reality, this is something we should have been doing from launch, but migrating from a legacy system to a new platform internally made this way more difficult than it needed to be—so until things were a little more stable, we had a page that was working.
Internally, there were people who were skeptics – would the ROI from A/B testing be a net positive?
The answer is yes. (which is no surprise to Conversion Rate Optimization specialists out there)
At PubCon in mid-march I saw a presentation from Chris Goward (writer of You Should Test That
) and some of the most valuable information I gleaned from it was that just following standard conversion optimization best practices wasn't enough—without a careful testing methodology you could potentially leave a huge chunk of conversions on the table.
So What Did We Test?
- Removing Distractions: This was one of the most important things that came out during our A/B testing process and just isn't something you could really learn algorithmically (without frankly being lucky). We took the standard design we had been using and we stripped out anything that may be a distraction. This things included: top of page navigation, non-essential footer information, secondary calls to action, etc. We focused the page down on the user doing the one task that mattered. 150% of our increase came from this change alone!
- Font Size and Fontface: At first we went bigger—pushing our standard paragraph text from a size 14 to a size 16 across the entire site. This yielded not statistically significant difference between the baseline. Afterwards, we moved away from the standard web fonts and leveraged a custom font. We tried this at both 14 and 16. The move to a more elegant font at a higher point accounted for nearly 20% of our Conversion Rate uptick.
- Responsive Design: The site had already, technically, had a mobile friendly design but it wasn't something we had put much stock into.
What a mistake that was.
By cleaning up the mobile experience we saw an increase of nearly 100% in our conversion rate. Just think about that and how stupid we felt! Making a strong user experience on the phone completely transformed our landing page and drove a dramatic increase in ROI.
- Everything Else: I’m saying everything else because there were a lot of small tweaks that in themselves aren't worth talking about…but the majority of it comes down to better copywriting –clearing up the user intent, making sure the text promotes the content and the brand both in tone, voice, language, and visual presentation.
What Are We Testing Next?
All that said, 300% is good but I think it is a drop in the bucket. I believe that with continued testing we could see a 1000% or more increase in our RPM from the baseline, and when we finally hit that mark, we've document the entire process in more detail.
So what key things did you do to see the greatest increase in your landing page conversion rates?
- Increasing Mobile User Experience with Retina-Display Ready Images
- Testing Entirely New Landing Page Concepts
- Greater Focus on Typography and Copywriting
- Increase Landing Page Diversification For Different Products and Services