Create Your First A/B Test

Crazy Egg makes it fast and easy to ship experiments that help you make a better site. We use the multi-armed bandit testing approach so that you can maximize your ROI. 

Before you  set-up an A/B test, you'll need to plan it out: decide where the experiment runs, use the Visual Editor in Crazy Egg to design your variants, and set-up a goal. Once you're ready, you can publish your experiment! Let's go through each step together.

Create a hypothesis

Before you can start testing, you need to start with a hypothesis. For example, a hypothesis can be, “If I change the copy of the “Sign-up now” button, the number of sign-ups will increase by X.”

To help you brainstorm ideas, check out the  Examples of Tests article.


Create an A/B Test

1. Go to  A/B Testing.

2. Click on  Create your first A/B Test or the plus (+) at the top of the dashboard.

3. Choose the page you want to test. 

4. Choose the audience that you want to test: (1) Desktop visitors, (2) Mobile visitors, or (3) Tablet visitors. 


Create a Variant

Next, use the Crazy Egg Visual Editor to create variants.

1. Click  Create New Variant to create one or more variants of your page that you'd like to test. Within the sidebar menu, the Control is always your original page. You CANNOT change the Control. Variant #1 on the hand is the adjusted version. 


Keyboard Shortcuts

  • Undo for Mac: cmd+z
  • Redo for Mac: cmd+shift+z
  • Undo for Windows: ctrl-z
  • Redo for Windows: ctrl-y

2. Once the Visual Editor has loaded your page, select an element that you want to edit. For example, let's select the  BUY T NOW - $10 button and change the copy.

3. Click on  Edit Text & Style

4. Use the inline editor to change the text. Click on the  close button (X) once you are done. Here, we've changed the copy from Buy Now - $15 to Buy Now.

5. Repeat steps #1-4 to create more variants
6. Click  Next once you have finished creating your variants.


Deleting a Variant

  1. To delete a variant, click on the variants dropdown. 
  2. Find the variant that you want to delete. 
  3. Click on the trash icon
  4. A modal will show up, click on Delete permanently.


Configure or Edit goal

Define the goal to measure the impact of your variants. There are four types of goals:

  1. Sell more products
  2. Get more registrations
  3. Generate more leads
  4. Get more pageviews

Once you've chosen a goal, you'll need to define how this goal is achieved. There are four ways a goal is reached:

  1. When someone clicks a button/link. For example, Crazy Egg will record that you sold a product when a customer clicks on the Buy Now!button.
  2. When someone reaches a page. For example, Crazy Egg will record that you sold a product when a customer reaches the Order confirmed! page.
  3. When someone submits a form. For example, Crazy Egg will record that you received a registration when a customer clicks on the Register button within a form.
  4. When someone triggers a script. This method is for more advanced, technical users.

In this test, we want to see if changing the text of the button from  Buy Now - $10 to Buy Now will get more people buy the product. So we've set-up the goal as Sell more products.

To determine this, I consider a product sold when someone has clicked on the  Buy Now button. A red star highlights the button I have chosen to track.

Set-up a name for this goal. Click  Next when you are happy with the goal you've set-up.


Name Your A/B Test

Once you have defined your goal, create a custom name for your test.

Set your test live!

Once you're done, click on  Start testing my ideas!

Important Note: You will  NOT be able to edit your variants once the test has started. Review everything before starting the test.

Congratulations! You’ve launched your experiment and now you’re ready to check out the Results page. It will take about 24 hours for your site to see your variants live.


Saved Goals

Once you set-up and save a goal, we save them for you for future use. Saved goals will only show-up for the domain and audience in which the original test was set-up. 

For example, if you created a goal for www.example.com targeting desktop visitors, this goal will not be available for www.example.com for mobile devices. This saved goal would only be available for www.example.com for desktop visitors. 

Still need help? Contact Us Contact Us