Make measurable experiments is one of the main advantages of Digital Marketing . After all, it is through them that we can find possibilities for improvement and optimize the actions that you already do.
For this, a known method is the A / B test , which is to create two versions for emails, landing pages and other parts: the first is the original version; The second, is more challenging with modifications. Then they are shown for the same number of people and compared. Finally, it is clear which is more efficient.
The A / B tests are excellent tools because they offer a real feedback from the market. There is a simple search on someone can answer one thing and doing another in practice: they are facts.
About what is A / B testing and how to make them, you can read more details in our Test Guide A / B . In this post, the focus is to show some examples we do here at Digital Results and found useful to be applied to make your Digital Marketing actions even more efficient.
Remember that, although some of them have failed, the result can be totally different with your company. So the best way to verify that an experiment will work for your business is … testing it!
Also remember to use the Test Calculator A / B Digital results to calculate the sample size required and check the test results.
The idea of this experiment was to change the color of the form of the free trial page conversion rate of the Landing Page.
We believed that by increasing the form contrast with the background, the form would have more prominence, improving conversions by 20% relative to the original Landing Page.
Doubled, then the original page, we have made changes in the second version of the page and begin the test. Finally, we evaluate the performance of the experiment, concluding that the challenging version (form with more contrast) had a conversion rate 27% higher than the original version (form with little contrast), confirming the initial hypothesis of the experiment.
In this test, included a Call-to-Action (CTA) to download the material at the beginning of the email body seeking to increase the click through rate of the campaign – which until then, with the button at the end of the text, ranged between 15 and 20% . The idea was that the click rate increased by 20%.
We create a new template, placing the CTA at the beginning. We shoot the emails and verify that the version with CTA at the beginning had a click-through rate 24.28% higher than the original version, confirming the initial hypothesis of the experiment.
We believed that indicate on a Landing Page the amount of downloads of the material that had been made could work as social proof, increasing conversions by 10%.
So we created a version of these pages with a counter downloads. The test was performed in Pages Landing Digital Marketing Kit for Beginners and eBook Digital Marketing for Entrepreneurs .
Add the number of downloads that stuff ever had and we set the A / B testing via RD Station Marketing.
In Digital Marketing Kit for Beginners, after 49 days of data analyzed, the original version of this kit had a conversion rate of 30.64%, while challenging version had a conversion rate of 31.25%.
In the case of eBook Digital Marketing for Entrepreneurs , the original version had a conversion rate of 29.67% (189 conversions to 637 visitors), and challenging version had a conversion rate of 30.64% (106 conversions to 346 visitors ).
However, the test calculator / B Results of the Digital considered not significant difference, and the test was inconclusive.
The idea here was to update the layout in Landing Page chat to increase the conversion rate. By changing the page template, we wanted the conversion rate had a 30% increase compared to the previous page.
The conversion rate of the original Landing Page stood at 11.46%, while the new Landing Page stood at 12.58%, a variation of 9.77%.
We therefore consider that the experiment failed because it did not prove the hypothesis that the conversion rate would increase by 30% with the change of layout. The main hypothesis for the failure was the instability of exchange rates, as they vary greatly from each other during the trial period, which may mean that, depending on the channel, a type of Landing Page can have better performance than others.
The main learning was that even improving the design of a Landing Page, not necessarily the conversion rate will increase because it depends on other variables such as the origin of the Lead .
We tested the inclusion in the Landing Page, a video of the author of an educational material ( eBook , webinar , etc. spreadsheet) making a call to the user to download the material.
When we change the image of the Landing Page by an author video talking about the educational material and inviting to download the offer, we believed that the visitor would feel closer and that the conversion rate would increase by 20% compared to the Landing Page that had only the image.
We chose the material, created the script of the video, which was then rehearsed, shot and edited. Finally, we have created a Landing Page challenging and begin to monitor conversion rates and return. Ufa!
However, the result was not expected: at the end of the experiment, we saw that the video can decrease the conversion rate of a Landing Page. Among the hypotheses for this to happen are distraction and people who entered the Landing Page with an expectation, and when they saw the video, did not have their expectations met.
Among the suggestions to do a new experiment, they are rewriting it with another material and put the video on the thank you page.
If you are looking more theoretical knowledge about A / B testing and on growth hacking , which is also a way to further improve your results, we recommend to supplement your learning by downloading our free Kit Growth Hacking and A / B testing .
And you, what experiments have right and wrong in your company? Share your learning with us here in the comments and good reading!