2024 Landing Page Test Trends

Optimizely, a digital experience marketing platform widely recognized for its website testing, is making waves with its recent endeavors. The company has been doing roadshows across the country, sharing fresh product updates, insights on integrating generative AI into its platform, and findings from its experimentation trends report. 

This report offers invaluable insights, as it is derived from analyzing over 127,000 landing page tests conducted within their platform over the past year. Collective Measures had the privilege of attending Optimizely’s stop in Minneapolis, gaining firsthand access to these insights. What did we learn? In this blog post, we’ll explore key takeaways from their report, provide our expertise on the insights, and share how advertisers can apply this knowledge to enhance their testing strategies in 2024 and beyond.

Proactive versus reactive testing

More often than not, landing page testing comes only as an afterthought for advertisers upon seeing that KPIs are not in line with what they would expect. For example, conversion rates are below average, Google’s landing page experience is “Poor,” or bounce rate is higher than normal. Only after seeing these types of data points do advertisers start to think about how they can improve their landing pages via testing.

This reactive approach to landing page testing is common, but it doesn’t have to be. Collective Measures encourages advertisers to think about testing before new pages are published. For example, incorporating testing plans into the launch cycle of a new campaign, product launch, etc. will ensure you hit the ground running with an approach to measure which page experience resonates best with your website visitors. 

A few ways this could look:

• Identify multiple relevant URLs to test in a new paid media campaign
• Test different page layouts before publishing a new page
• Rotate different hero content on your page to see which drives more engagement

Adopting this proactive approach will save marketers time as they no longer have to wait to learn that benchmarks aren’t being met to take a different approach. As advertisers, we don’t always know what is going to work. Testing eliminates the guesswork and validates which outcome will be more impactful.

Multivariate tests > A/B tests

Within Optimizely’s data set, they show that experiments that test multiple treatments are 3x more successful than A/B tests. Their reasoning? More complex tests stray away from the more “safe” options in favor of trying out original ideas that can move the needle more quickly. Minute tweaks have minute effects. To really impact user behavior, marketers should be implementing tests that rethink the journey the customer takes from click-through to conversion.

You win some a few, you lose some a lot

Optimizely has reported a 12% win rate for experiments run on their platform. This means that 88% of tests either do not win against the treatment variation or do not reach statistical significance. This may seem low, but it lines up with other reported testing data from industry leaders such as Google, Amazon, and Airbnb.

The point is to not be discouraged if you implement a test and it fails. Because you completed the test, you now know what worked and what you only assumed was working. There are still valuable insights to be gained from an inconclusive or failed test, including uncovering areas that may not be of interest to your audience or helping inform you of tests to avoid in the future.

Test across the whole funnel

Understandably, a large percentage of landing page tests focus on improving primary metrics such as revenue or conversion volume. While these tests should absolutely be conducted, a big miss in the industry is not taking time to consider testing other elements that are conducive to achieving the larger goal. Examples of this are improving the menu or navigation experience, enhancing search bar performance, and improving the engaged session rate. 

According to Optimizely, users who perform a search on a website are 2x–3x more likely to convert than users who do not. It is important to acknowledge the impact that these features have on a user’s ability to reach the end goal in the first place, so they should definitely be incorporated into our testing plans.

What marketers need to know

As the search landscape continues to evolve in the age of generative AI, machine learning, and large language models (LLMs), landing page testing needs to be incorporated into advertisers’ plans from the onset. This ensures that advertisers understand what resonates with their audience earlier than ever. But don’t forget about current traffic. Marketers still need to optimize for and retain the traffic they DO get to their websites given the magnitude of changes disrupting the search landscape as we know it. Incorporating the above thoughts into your testing plan in 2024 will lead to valuable insights marketers can continue to build on for years to come.