How different privacy policies affect conversion: 4 A / B test results!

When creating forms for collecting contacts or for registration, it is important to test different options in order to understand which works most effectively. In this case , written in the first person, by Michael Aagaard , an example of how you can, only slightly changing a small microcopy on a form, change the conversion rate. A microcopy of the company's privacy policy can be written in different ways, and as the author discovered in this case, the result will also be different. How to write about the privacy policy on the form so that as many visitors as possible leave their contacts and register?

Recently, I tested 4 different options for using the privacy policy on the registration form on the home page of the community of betting and betting community. The results surprised me a lot, since each of the options had a different impact on the number of registrations - from a decrease in registrations by 18.7% to an increase by 19.47%.
')
In this article I will show all four variations, introduce you to the data obtained as a result of testing, and share what I learned from each of the experiments.

The first experiment with privacy policy

Here is the screen of the main page of the site BettingExpert.com. As you can see, the initial version of the registration form does not include a privacy policy element, so I decided to check how the conversion (in this case, the number of registrations) changes if I add a privacy policy to the form.



In my first experiment, I preferred the informal, unobtrusive privacy policy option:

100% confidential - we will never send spam

I was absolutely sure that using the privacy policy would have a positive effect on the number of registrations compared to the original form. In fact, I conducted testing only to find out how big the jump in conversion rate will be. Imagine now how much I was puzzled by the test results!

A form with an emphasis on maintaining confidentiality led to a fall in conversion and reduced the number of registrations to a staggering 18.7%!



Test data: I tested for 9 days, 16152 visits and 297 conversions (or form fillings) were carried out, and the initial version was more effective than the one that focused on confidentiality throughout the trial period. The statistical level of data reliability was 96%, and the standard error was 0%.



The development of the first experiment over a 9-day period.

What I learned from the first experiment:

No matter how paradoxical it may seem, adding a privacy policy does not guarantee more registrations. On the contrary, it can reduce the conversion.

My hypothesis is that, although in the message you are trying to reassure potential customers that you will not be spamming, the word “spam” is disturbing in the minds of users. Therefore, directly in the form of the word itself should be avoided.

Second experiment with privacy policy

In the second experiment, I did not repeat the mistakes using the word “spam”, but I still used a short and alluring statement about 100% confidentiality, as in the first experiment:

100% confidentiality. We will keep all of your personal information private.

Test data: I conducted this testing for 12 days , and after 15675 visitors and 279 conversions (registrations), I did not notice a significant difference between the original version and the second experimental form. In the first days of testing, the effectiveness of the second variant was higher, but closer to the 12th day, the results were equal.



At first, the progress of the second form was noticeable, but closer to the 12th day, the results were equal.

What I learned from the second experiment:

Despite the fact that the second experiment was more successful than the first, it still did not lead to an increase in conversion.

It would seem that the fact that I removed the word “spam” and focused on the aspect that the information would be kept secret should have had a positive effect. However, even with a focus on keeping information secret, the company's privacy policy still seemed to be vague and uncertain to potential customers.

The third experiment with privacy policy

The first two experiments helped me to get really valuable information and contributed to the idea for the third experimental registration form. By the way, I was still surprised by the results of the two previous tests.

For the third time, I decided to make the privacy policy more authoritative and firm , instead of gently pointing to keeping information secret. In addition, I wanted to make this option much more understandable than the previous two. So, having thought it all over, I ended up with:

We guarantee 100% confidentiality. Your details will not be distributed.

Bingo! This option worked! He significantly increased the number of registrations - they have increased by 19.47%!



Test data: I conducted this test for 12 days , 20,257 visits and 380 conversions (registrations) were recorded . The statistical level of data reliability was 96%, and the standard error was 0%. From the very beginning of testing, the effectiveness of the third experimental form was higher than the original version.



The course of the third experiment during the 12-day testing period.

What I learned from the third experiment:

The guarantees provided were the most notable change in the third experiment. The first two talked only about 100% confidentiality, while the third experiment was about guaranteeing 100% confidentiality.

It seems to me that this formulation is much more credible. In addition, the second part of the message, which says that your data will not be distributed, looks not only understandable, but also much more convincing.

Persuasiveness, clarity and credibility - these are the parameters that should be taken into account when developing a privacy policy, and I believe that a combination of these three factors is exactly what the success of the third experiment is connected with.

Fourth Privacy Policy Experiment

The first three experiments taught me really a lot and helped develop several hypotheses - customers need a guarantee of your privacy policy; the use of the word "spam" can seriously harm.

Now I was curious to find out what would happen if I combined the “best” and “worst” elements of the previous experiments, connecting them into one - this is how the 4th experiment was born:

We guarantee 100% confidentiality. We will never spam!
Test data: I conducted this test for 15 days , 18,959 visitors and 370 registrations were recorded , and as a result , no significant difference was found between the original version and 4 experiments.



What I learned from the fourth experiment:

The fourth experiment was a combination of the best element “We guarantee 100% confidentiality” and the worst “We will not spam!”.

The data obtained as a result of testing showed that there was no significant difference between the original version and the experiment, the “good” and “bad” parts somehow neutralized each other, so there was no noticeable effect on the number of registrations.

The results of four experiments:

The privacy policy that you use in the registration form can have a big impact on the registration rate. However, the use of the privacy policy does not guarantee a higher conversion - moreover, if you do not take care of choosing the right wording, the conversion rate may drop significantly.



But if you are willing to spend time exploring what really “works” in the case of your particular form, a privacy policy can lead to a significant jump in conversion.

The results of my research suggest that a trustworthy, understandable privacy policy, guaranteeing the security of the data of potential customers, will help them feel secure when filling out the form. Moreover, the findings suggest that you should be careful with the word "spam" - even if you are talking about its absence - as this may contribute to the opposite effect and may cause users to feel anxious.

In this series of tests, the most effective privacy policy turned out to be the following:

We guarantee 100% confidentiality. Your details will not be distributed.

Now I am testing this option in my subscription form to the mailing list on ContentVerve.com, and at the moment the updated version is 6% more efficient than the original version - the data is still inconclusive, but still.

By the way, this is not the first thing I test on BettingExpert.com.

I have done many other tests on this site. Among other things, I tested several versions of the text, which led to an increase in the number of registrations by 31.54%.



PS This case interested us, the online designer of the collection of contacts, promotions, social sharing buttons and other tools, including because one of our subscription form templates looks like this:



The client who edits this template for himself can write some other note, we hope that this article will help you choose the right privacy policy. After reading this article, we also decided to test different text on our subscription forms with the message that user data will not be distributed on one of our clients' sites. What came out of this, we will tell in one of the following materials.