Website A/B Testing: 4 tips to beat an unbeatable landing page
Have you ever had a landing page you just couldn’t beat?
I know I have.
I tried everything I could think of, and yet, one treatment after another was beaten by the control.
Today’s MarketingExperiments blog post features four tips I discovered for beating a reigning champion that you can use to get your testing mojo back. Our goal is to give marketers a few insights to aid testing efforts.
But before we get started, a quick caveat. While increasing conversion rates is important, you shouldn’t waste too many resources going after another lift unless the ROI on your efforts makes sense.
Tip #1. Dive deeper into your data for new insights into your customers
Things change …
… and so does customer behavior.
Consequently, taking the time to reassess your data and tracking efforts can help you discover if any changes in traffic behavior have occurred during the time the control has been unbeatable.
One good example of digging deeper for new or alternative data is enhanced click tracking.
We discovered through testing with one Research Partner, 20% to 30% of the traffic being driven to our control landing page was from people who were already users of the product being offered and had almost no chance of converting.
These visitors were being redirected into the sales funnel from a source page we did not have any control over. By implementing click tracking on a specific link on the page, we were able to identify the problem and direct these visitors to the appropriate section of the Partner’s website.
Ultimately, we were unable to update the Partner’s source page, but we were able to assess an adjusted conversion rate and design tests to better cater to prospects versus existing users of the product that our analytics platform categorized as “new users.”
Had we not taken the time to learn more about why so many users were clicking away from the page to continue looking for services related to their already existing account, we may have never discovered a new opportunity to challenge the existing control.
To further aid your efforts, you should also consider additional tools to generate alternative metrics that can help you learn more about your audience. Here are a few of my favorites:
- Scroll tracking – Alternative metrics based on how visitors use scrolling on your page. You can use this to learn more about how visitors interact with your content.
- Heat mapping – Visualize customer behavior on your site by using heat mapping tools to learn more about where users focus their attention on your site.
- Form field error tracking – One of the most underutilized analytical tools out there is form field error tracking. This allows you track errors that occur in the form field process you can use to identify any potential friction or anxiety that may exist in your form process.
The bottom line here is taking a deeper dive into your pre-existing analytics and tracking while adding some new alternatives into the mix can help you develop new test ideas built on fresh data.
Tip #2. Look to your peers for different ideas
I remember once staring at a landing page so much that instead of seeing opportunities for conversion, I instead saw database fields and possible back-end improvements that were unlikely to produce any significant results or discoveries.
That’s when I realized it was time to take the unbeatable landing page to one of our peer review sessions to hear some fresh perspectives.
The big idea here is that it’s easy to get so familiar with a landing page or process that you can miss new testing opportunities if you’re not careful. Therefore, getting a little help from peers or a trusted colleague can pave the way to discovering the more elusive test treatments that can really aid your efforts to beat the control.
Tip #3. Look to different channels and platforms for answers
Most of the time, one control alone can’t outperform treatments across every sales channels, not including the myriad of devices that exist.
In short – what works on a landing page may not bode well in the mobile space.
This is where tracking test performance at the channel and device levels can open up new windows for treatments to test against a stubborn control. For example, a treatment that produces a multistep process focused on reducing cart abandonment may work best in a desktop browser.
For mobile or tablets, however, a single-step process that utilizes screen sizing and swipe-friendly technology may be ideal.
That may seem obvious, but according to the MarketingSherpa 2012 Mobile Marketing Benchmark Report (free excerpt at that link), only 13% of marketers routinely review tests and decide on a test plan for follow-up testing.
This means opportunities for capitalizing on low-hanging fruit in the mobile and tablet spaces can also lead to discovering new opportunities for treatments that derive from a multi-channel marketing perspective.
Tip #4. Does your control fully meet expectations set by ads
Each channel drives traffic to your landing page with different motivations. So, taking a step back to ask yourself if the control is meeting customer expectations could be the key insight necessary to develop an effective treatment.
One classic example of this is how PPC ads set different user expectations for a landing page versus visitors coming from organic search. The common mistake marketers make here, however, is placing the focus solely on landing page optimization and not testing the ad.
This results in an optimal landing page left to the mercy – and underperformance – of an untested PPC ad.
Ultimately, don’t be afraid to try new things, but make sure your decisions are well-informed, backed by data and generated from a position of true customer insight.
Do you have any tips for beating that unbeatable control? I’d love to hear your thoughts in the comments section below.