We have convinced you how important doing A/B tests for every landing page is and how it can contribute to its suitable optimization a few times so far. Testing the effectiveness of particular elements on the landing page will undoubtedly translate into overall results of the campaign. We have written, among others, about what elements ought to be tested and how we should do it. Read more about it in the articles: A/B tests – website testing, or Test forms on the landing page! Now it’s time to briefly remind about some general rules concerning appropriate landing page A/B testing.
What are A/B tests – a quick reminder
A/B tests consist in creating two variants of the landing page differing in one element only. In A/B tests one should not test a few elements at the same time, as one will be unable to examine which one of them had influence on greater effectiveness of the landing page. Such kind of a test will allow for specifying which element on the landing page should be changed.
What to remember when doing A/B tests?
1) Test simultaneously
What does it mean to test landing pages simultaneously? One needs to remember to test two prepared variants of the landing page at the same time. In this way any factors that could disturb the results will be minimized. Elements that could negatively influence the reception are numerous. First of all, one needs to take into consideration seasonality, that is the fact that customers behave differently on weekdays, during weekends as well as during holidays or vacations. The correct course of the tests should focus on showing two versions of the landing pages interchangeably, that is one net surfer is shown variant A and the other one is at the same time shown variant B, etc. If you act in a different way and show each of the variants on a different day of the week, for instance, you are sure to receive distorted results.
2) Test for at least two weeks
Testing of the landing page should not last for too short. It is about achieving the most representative sample and it will be possible only over a longer period of time. Measurable results can be obtained by testing the landing page for at least 10-14 days.
3) Ensure at least 100 conversions/1,000 visits
A long time of duration of the campaign doesn’t guarantee the appropriate number of conversions. Make sure that the test achieves a sufficient number of actions: it will allow you to receive more detailed results. You should gather at least 100 conversions or 1,000 views, so that a more efficient version of the landing page can be determined.
4) Test the move on new customers
It is good to think about which customers will be an appropriate group that we can test the landing page on. Those that we already have or maybe should we target the test at a different group? The present customers may not be appropriate, as their habits might distort the results and lead to inaccuracies. Doing a test on a new group that hasn’t had a contact with your landing page is a definitely better choice.
5) Test constantly
You’ve finished one test, you’re happy because you know what to change on your landing page and now you’re waiting for correction of the results of the campaign. However, in order to strive for the best results one should not stop at one test only. Therefore, test, and then test, and after you’ve finished… run another test!;) It’s all for the sake of further improvements of your landing page, pursuing the set aim and satisfaction from the campaign.
Before you begin to do tests, analyze the statistics of the page up till now. Check what the bounce rate and the conversion rate are and how much time the users have spent on your website. Such analysis will allow you to state precisely the goal you want to reach after running an A/B test of the landing page. Treat this constant testing of the landing page like a fight for the results you care about.
If you like our article and would like to create the landing page – sign up for free and try Landingi app.