A/B testing is an absolute must-have for companies who want to optimize conversion rates of their websites. It’s not difficult anymore and you can use one of the popular (and sometimes free) tools like Optimizely or Convert. You’ll be able to prepare two or more versions of a page to see which one performs better. But that’s not what this article is about. If you want to know more about conducting A/B tests, take a look at these comprehensive guides from Fitsmallbusiness and ConversionXL. I’m going to talk about using feedback from visitors to collect ideas for A/B tests.

Why? Because it’s easy to get stuck with coming up with new A/B testing ideas. There are dozens of guides with best practices of running A/B tests. The most popular ideas are changing colors of buttons, adding scarcity factor, simplifying purchase path etc. But there are many more elements worth testing. Analytics might help you (e.g. find pages with the highest exit or bounce rate, see where on a purchase path users drop etc.) but there are many possible tests to run on pages which report average stats. How to discover those ideas?

Collecting opinions of visitors will help you prepare new and clever A/B tests that otherwise wouldn’t have been run. If you’re wondering how to collect feedback from to get A/B testing ideas, I recommend using website surveys – you will quickly collect a meaningful amount of data and targeting options allow you to precisely target who and when sees certain questions.

In this article, I’ll show 5 questions to ask your visitors to get their ideas for A/B tests.

What do you like the most about this page?

At a first glance, this question is not the right to ask when collecting ideas for A/B tests. But keep in mind that people don’t like big and sudden changes. Before you start testing, it might be a good idea to find out what people like the most on your website and not to change these elements, especially if a big portion of your traffic are returning visitors. If you want to test changing them anyway, a gradual change might be a better idea. Changing everything at once can distract returning visitors.

What do you dislike the most about this page?

That’s probably the most basic use of feedback to get A/B testing ideas. Simply ask visitors what they dislike about your website or a certain page. Then spot the most common answers and turn them into a variation of a page. Sometimes profound changes in a good faith like updating the design to make it look modern can lead to bad results if they include factors that users like the most. Learn more about such failures here: http://www.sitepoint.com/3-painful-site-redesign-disasters/

What would you change on this website?

Being direct is the key. Simply ask people what they would change. It’s a bit different that the previous question about elements people dislike. There might be elements that people generally like (or don’t mind) but have ideas how to change them. Here you can expect more creative ideas and you’ll see how innovative your visitors are since answering this question requires more effort than the previous one.

What is missing on this page?

A/B tests are not only about changing buttons or rearranging elements. It’s also about website content and other tests whose goal is to convince more visitors to buy. Providing all information visitors need is one of the crucial factors to make them more likely to buy. Stats show that surprising delivery costs are one of the most popular factors why people abandon shopping. So maybe it’s worth to test adding this piece of information to products pages if you’re not doing it yet?

Why didn’t you buy/register?

This question will provide you with more general feedback. Probably you’ll find out that some people find your products too expensive. But some people might decide not to buy due to factors easy to change and test. Example: people drop off because they aren’t sure whether you provide their preferred payment method or can’t find any security badges. You can fairly easy change it and see what results it brings.

Text answer question - Segment

Tricks to get more a/b testing ideas

If you decide to use Survicate to find A/B testing ideas you can do the following trick: tag people who answer in a certain way. Example: you ask a question what visitors would like to change and you give them 5 options. Tag each respondent. When you decide to implement one of the suggested changes run a survey targeted at visitors who suggested it and ask how they assess the change. Learn more about targeting based on tagging here.

Another trick: ask returning visitors how they assess changes – they are likely to know the previous version so they can compare. Don’t make a mistake of asking this question to all visitors – new visitors will have no idea what you are asking about and you’ll look foolish. Take advantage of targeting options to make sure your surveys are well targeted and provide you with useful information without disturbing or surprising visitors.


As you can see, your visitors can easily become a great source of A/B testing ideas. Just remember about sample size – when one person says that you should change the color of your website or slash prices by 90% shouldn’t be enough to start a test. Wait for more answers and turn the most popular choices into A/B tests.