skip to Main Content
A/B Testing- You’re Doing It Wrong!

A/B Testing- You’re Doing It Wrong!

I?ll jump right in. The way you?re designing your websites- it?s all wrong. You?re also not testing it right. The whole process should be reversed.

Instead of designing your website on your own, let your customers design it for you. Don?t try to create a Minimum Viable Product (in this case website) and only then let your customers access it. Let them join the design process early and immediately create a product they are sure to like.

Similarly with A/B testing. Don?t spend weeks on coming up with a perfect test. Start testing right away and EXPERIMENT! If you want your A/B tests to have a significant impact on your conversion rates, then you will have to test big changes. Test changes that are bold and brave and big. There?s a reason A/B tests are often called ?experiments? by CRO experts- because they allow you to experiment with your website and business!

Don?t change the color of a button and call it a test. Turn your whole website upside down- delete that header, add pop-ups. Delete any copy on your landing page and put a video there instead.

Don?t hesitate when deciding what to test and waste time. Don?t spend days and days coming up with that test idea, but go with your Riskiest Assumption. That change to your website you think is too radical? That?s the one you want to test. You want each test to be a Riskiest Assumption Test.

Welcome to the age of the RAT!

The problem with MVPs
RAT vs. MVP

Understandably, you want everything your company does to be well executed, thought out and perfect. That?s why a lot of companies wait until they have their Minimum Viable Product to put it on the market. The problem here is that when you build an MVP, whether it?s a website or merchandise, you build something that you like. Better said, that your conscious mind likes, and the conscious mind is NOT what drives conversions.

Consumer psychologists have known for a long time that it?s the subconscious that you need to appeal to, if you want to increase your profits.

In order to design a website that will bring you a lot of revenue, you have to create a website that will appeal to the subconscious mind of your customers. This means you need to test them consistently to see what they want, because by constantly testing their behavior you will see what they like. Otherwise, if you ask them directly, they?ll tell you what they think they want. But this is the conscious mind talking, the part that doesn?t have a lot of control and doesn?t drive profit.

Some of the biggest companies on the planet understand this and have based their growth strategy on this idea– appealing to their consumer?s subconscious.

A lot of very famous websites would look different if they had designed them according to what their customers ??said?? they wanted. Customers will usually tell you that they hate pop-ups and lots of things flashing on the screen. They want a sleek and minimalistic experience. But it turns out that pop-ups and other elements that ??urge?? the customer to do something, really work. They influence the person?s subconscious and can help you increase your sales.

The issue with the MVP is that, when creating it, you?re creating something you like and what you think your customers will like. And then when your customers don?t actually engage with it, you start tweaking it until you get to the point where it?s something that their subconscious mind loves. You could?ve jumped in straight into letting your clients design your product!

Why not save yourself some time and immediately create a product that your customers will love? Don?t waste your time on MVPs.

Riskiest Assumption Test

A lot of companies apply the same logic to their A/B testing strategy. They spend weeks and weeks on research and analyses, work on the design, and only when they think they have an MVP test- they run it.

Even worse, when coming up with the test idea, they waste a lot of time creating ??hypotheses??. The problem with ??hypotheses?? is that they?re still just assumptions, regardless of the work that has gone into them.

Most ??hypotheses?? aren?t based on real data. A lot companies think they have credible data and build tests on that, but the truth is that you need a lot of tests and monitoring to come to a point where you have valid data. If you?ve done thousands of tests and have a sample of tens of thousands of clients- that?s when you have enough data. Otherwise, if you?ve only been doing 30 tests per year, that?s not enough for you to figure out how your clients behave.

So you?ll spend weeks working on a hypothesis, only to end up with what is essentially an ??assumption?? you could?ve tested immediately. That?s why you should embrace the idea that you?re working with assumptions right away. Not only because of the lack of data that prevents you from crafting a hypothesis, but assumptions are the fastest way for you to see an increase in profits.

Test big changes

A/B tests that experiment with small changes are not going to have a big impact on your business. When designing an A/B test, don?t be afraid to disrupt what you?ve been doing so far. Assume that the riskiest change is going to bring you the biggest increase in conversions.

Don?t be afraid to completely change what your website looks like. Don?t just change the color of a button or some text in your header. Turn things upside down. Change the design of your website if you can and test how your customers will react to that. Add a landing page with big images if you don?t have one. Completely change your tone of voice and message. Summarize your mission and vision into an illustration. Add pop-ups and motivators.

Go for the riskiest test idea that comes to mind and run it. And then quickly jump to the next Riskiest Assumption Test, not wasting any time on your way to growth.

Conclusion: Test fast and constantly

Moving away from the MVP model, both when creating your website and designing your A/B tests, enables you to do more tests, faster. Even if you?re doing 10 tests per year, the RAT model enables you to quickly jump from one idea to the other.?This is especially relevant when you have limited resources and can?t afford to spend time building an MVP test or website, but want to start testing right away.

Your testing strategy should always be lean and agile, because the more tests you run, the greater your chance of seeing a significant increase in revenue. By constantly tweaking your website, you have an opportunity to cater to your customers better. This way you?re going to be sure that you?re creating a product modeled on what your clients actually want and not what you think they want.

Just ask Booking.com! They themselves say that ??everything, absolutely everything is tested?? on their website and at any given moment they run around 1000 A/B tests.

You don?t want to waste a lot of resources when preparing tests and want to go through them quickly. That?s why modeling your testing strategy on RAT is the perfect solution.

Don?t have a big team at your disposal that can develop all the A/B tests you need??Team Croco?can help. Our specialty is fast test development and there?s no test we can?t code! If you think we can help you, reach out to our team?here?and we?ll answer all your questions.

Written by Lela Dragani?. Lela is our Content Manager and can be found rummaging in the depths of A/B testing theory. When she’s not behind her laptop, you can find her packing her suitcase and looking up directions in Google Maps (they rarely help and she’ll still go down the wrong street, but oh well- adventure!).

Back To Top