5 A/B Test Ideas For Better Landing Pages
Updated: Jan 14, 2019
Many places on the internet will claim they've discovered the secret to higher conversion rates. And for just your name, contact email, firstborn son, three payments of £19.99 and phone number, they'll happily supply you with all of the answers.
The truth is there is no magic secret to getting more conversions. The success of a web page depends on a combination of hundreds of factors. What worked for one brand or campaign may have the completely opposite effect for another!
So we're not going to tell you how to get more conversions. Instead, we thought it would be better to share some of our findings with the hope that you test them yourselves! Let us know if you have similar results or find something completely different.
Here's 5 A/B test ideas for you to try and (hopefully) achieve higher conversions on your landing pages...
1. Declutter above the fold
Just in case you're unsure what we mean by "above the fold", it refers to the average length of a web page that people are shown without scrolling. Due to differing screen sizes and browsers, this won't always be the same. Check out Hotjar for a good tool that shows the average fold of your users. This space is your prime real-estate, your shop window, your book cover, your opening line. People are going to judge you on it, and if it's not up to scratch they might not bother sticking around!
A busy webpage with lots of text information and a form on the right half of the screen wasn't converting very well on desktop. We deleted all of the text and centralised the form. The result? A 23% increase in conversion rate! With the user's entire focus on the form, more were inclined to click through to the end!
This principle seems to apply not just to elements on your web page, but to the basic design too. At one point, we had a header bar on most of our landing pages, with the brand logo displayed in the top left. Because that's how most web pages look right? But on a landing page without a navigation bar there is little need for one! We questioned why we had it, and couldn't find a good enough reason to keep it. With less noise and more focus on the form, conversions increased by 231%!
2. Test even when it seems obvious
What information to collect and how to structure questions is always up for debate. It's something that we think about a lot. We want to ensure that we're collecting every detail we need in an efficient way. A conversation about this very topic sparked our sales manager to make a suggestion...
"Why ask three questions in a row when we could get the same info in one?" She thought it was obvious that less questions = less time spent = more conversions. So we tested it. Three separate radio button questions became one checkbox question. Fewer questions, easier to fill out, more conversions - right? Wrong.
The short and simple version had a conversion rate of 9.77%. The longer journey? 17.69%.
The questions we cut were "fluffy" questions that were easy to answer. Their greater purpose is that once we get to the more difficult questions that people don't like answering (e.g. giving their phone number or email) they can see they're already halfway through the process, so might as well continue. They already feel involved and so won't give up so easily.
Often A/B tests surprise you by proving the opposite of your hypothesis, so don't take anything that seems 'obvious' for granted!
3. Try a popup
Putting pop-ups on our pages is something that's still a little new to us. After a couple of tests, we're learning fast and already have some findings!
There are a few different ways to go about it. The first way is through unbounce's 'Popups and Sticky Bars' function. It allows popups to show on arrival, after a specified delay, on exit, on scroll, or when something is clicked. Unbounce allows you to run an A/B test on what popup shows and can easily integrate with whatever you use to build your pages on.
We have run popups that show when a user moves to exit on a few different pages, with varying results. When used on desktop form pages, conversion rate jumped from 8.64 to 12.14! Though we had these amazing results on form pages, popups on advertorials actually had the opposite effect, bringing the clickthrough rate down on one page from 24.72% to 14.62%.
The second approach to popups is the social proof route. You'll have seen this used mostly on hotel and holiday booking sites, or gig tickets. They play on people's anxiety that the thing they want will sell out fast. We used ProveSource to show how many consultations were booked in for a treatment as people were going through a form page. Sadly, our conversion rate almost halved. We're putting this down to the campaign being unsuitable, but will definitely be giving this another go very soon.
4. Don't focus on conversion
Bear with us. We haven't gone crazy just yet.
Most landing page tests you run will have the end goal of improving conversion rates. But there are other factors that may be valuable to improve.
One less-than-obvious example is the quality score given to your ads on Google. The higher your score, the less it costs you to rank higher on search. If you don't know how bidding on Google works, then Google's Chief Economist gives a detailed (if slightly boring) explanation in this video.
Having high-quality scores can save your company an enormous amount of money. It's worth looking at Google's guidelines on what makes a good landing page and applying them to your pages. At the moment, we're working on improving a specific ad set's quality score with a dedicated landing page for that subset that includes easier navigation and more transparency around the aim of the page.
5. Tiny changes = big results
On first look, the two pages are exactly the same. But when users start exploring and scrolling down the page, the behaviour and responsiveness is very different. We didn't need to change what was on the page, and instead focused on how it behaved. This fix was simple to implement and the results were fast: an extra 3% on our click-through rate! Not bad for a few lines of code.
We hope you've got some inspiration for your next A/B tests. If you give any of these ideas a go, let us know! After all, sharing is caring.