Marketing Guides

What Should You Test?

What Should You Test?

As we mentioned previously, most A/B tests focus on elements that play some kind of role in conversions. This is why A/B testing is considered a major aspect of conversion rate optimization (or CRO) – that is, the practice of improving a website to boost conversions.

This doesn’t just limit you to changing the size and color of your “add to cart” button or the appearance of your “contact us” form, however. Let’s explore some of the elements on your site that might be perfect for testing.


If you use buttons anywhere on your site, these are an easy testing element to start with. Try changing the color, size, or even shape of your buttons to see if visitors respond any differently.

You may not expect switching the color of an “add to cart” button to have any noticeable effect on your conversions, but several tests have demonstrated that color can have a big impact on a visitor’s decision to click. In fact, this test detailed by HubSpot shows that a change from green to red boosted conversions for one site by 21%!

green and red buttons

Calls to Action

Along with calls to action that take the form of buttons, you should also think about testing text or link calls to action (CTAs). Using stronger or more targeted language can motivate an increased number of visitors to take the action you desire.

link to contact page

For example, if you end a page of content with a link to an additional page full of related information, you might test changing the color of the link, using a slightly different font, making the link longer, or just changing your wording.

Also, if you have a very important call to action, you might try testing it in a pop-up instead of simply placing it on the page. Pop-ups can be highly effective, but you’ll need to be careful to make sure they don’t drive conversions down by pushing your visitors away!

sample popup


If your website collects leads with a contact form, there’s a lot riding on that one form. Consider testing its length, number of fields, or appearance to see if you can collect more information.

One important thing to consider with forms: sometimes you can drive away potential leads simply by asking for too much information. You may want to test reducing the number of fields to a minimum, or adding clearer labels to indicate what is optional vs. required.

form example

In addition to contact forms, any form on your site can be tested as well. For example, if you run an ecommerce website, you can try a one-page checkout vs. a multi-page process to see what your visitors prefer. You can also test different labels on your checkout forms, pre-filling fields, and so on.


A little earlier, we gave the example of product copy being placed on the side of a page instead of below the product images. This is one example of a location-based test. Maybe a button on your page would perform better in a different position on your page, or a link would receive more clicks in a different place.

button comparison

You can also test the locations of elements like signup forms for your email list, or anything else that may be linked to conversions. It’s possible that an email CTA doesn’t have much effect at the top of your website, but converts extremely well at the bottom, where visitors arrive after reading your content.

Visual Elements

If you can change it, you can test it. The visual elements of your website that you should test include colors, fonts, layout, spacing, special effects, and much more.

An easy test would be to change the color of your site background. If you use a dark background, try testing a lighter one to see if visitors think the B version is easier on their eyes – and worth browsing a little longer.

Other Ideas

What else can you test? The sky’s the limit, as long as it’s an element you can change, and that you can effectively measure the results for after a set period of time has passed.

If your company has an ecommerce site, you may even consider doing A/B tests for your pricing. This may involve showing pricing packages differently, or even reducing the costs of some plans for a test group to see if more of them show interest.

button comparison

Chapter 1: Introduction to A/B Testing                Chapter 3: A/B Testing Software