How to Conduct a Successful Landing Page Test

Previously, Jana Fung of Mixrank discussed the four different types of landing page test and the importance of testing it. (You can view the entire webinar here.)

Now we will go through a checklist of variables that are required to conduct a successful landing page test. To conduct a successful test we must have a checklist to follow. This checklist is composed of procedures that is very, very similar to the process of scientific inquiry. This checklist is just one of many tips and tricks for conversion optimization. 

Requirements for a successful landing page test

There are different tests that can be done on a landing page, but each test has to have essential elements in order to achieve results that will ultimately be for the improvement of the landing page and the website as a whole. As discussed in this article by Unbounce, having a winning landing page is essential for a successful marketing campaign. The requirements for a successful landing page are:

1. Baseline
2. Make a hypothesis
3. Testing
4. Sample size
5. Simultaneous running

Setting a baseline

Before you start testing anything, you need to make sure that you are employing a web analytics tool that has been measuring your conversion rate on your existing landing page. This will give you a baseline of how your existing landing page has been performing. It doesn’t matter whether you have a low conversion rate or a very strong conversion rate. You have to make sure you know what the conversion rate was before you can start measuring any of the improvements.

Conception

Once you have an idea how your landing page has been performing historically, you can start taking on concepts that you think will increase your conversion. Once you have at least one hypothesis of what could be impacting your conversions you can start creating different versions of your landing page. But do not be limited by your own baseline, you should also have a point of comparison with your competitors.

The starting point of a marketing campaign revolves around your ability to define a point of differentiation. What is it about your product or service that sets it apart from the competition? You need to communicate this in a succinct way on your landing page. – Unbounce

Testing

As mentioned before in our blog, it would only require one additional version of your landing page that would allow you to test one variable on your page. Although with a multivariate test, you can have more than one additional version of your landing page. The only difference is, with multivariate test, you can have a lot of versions of the landing page, but that it will take a longer time to achieve a conclusion to find the best performing landing page.

Sample size

Next, make sure all your landing pages are taking the same amount of traffic. It would not make much sense to get majority of the traffic to one landing page and only test the others with just a handful of visitors. The reason behind this is that the volume wouldn’t be comparable. You could have a 50% conversion rate on a landing page with 1% traffic. But if you only have 2 visitors to that landing page and you get one conversion, it will cause the conversion rate to be 50%. It could be random and might not reflect the rest of the population or the rest of your target audience. So if you have two landing pages, make sure both are receiving 50% of the traffic. If you have three, make sure they each get 33% of incoming traffic. With four, give them 25% each, and so on.

Synchronize

Lastly, in order to make sure the landing page tests are done correctly, you have to run all the landing pages simultaneously. The audiences’ interest level is also a factor in conversion, and it could change from day to day, or month to month.

[Tweet “In A/B testing, testing different variables at different times will be irrelevant since the conditions will also be different each and every time”]

For example, if you are trying to drive umbrella sales on a landing page in the summer and then you want a different version of that landing page in the winter, you won’t be able to know exactly what design will increase your conversion since those are two different seasons. and obviously you will get a higher conversion rate in the winter because it rains more during that season. So you need to remember that all landing pages must be run simultaneously in order to draw a sound conclusion about the conversion rate and performance. If you want to learn more testing your landing page or you want to hear more about the webinar, you can view the entire presentation here.