How to use A/B testing for widgets | Wix

What is A/B Testing?

A/B testing allows you to compare two or more variations of a widget to determine which version performs better. By testing different designs, content, or offers, you can make data-driven decisions to improve customer engagement and conversion rates.

How does A/B testing work in our widgets?

A/B testing is automatically built into all widgets once they're set up. This feature allows you to create multiple variants of a widget and test them against each other.

For example, you can test:

Different headlines
Varying discount amounts
Alternate products

Setting up A/B testing

Follow these steps to enable and manage A/B testing for your widgets:

Create your first slot and widget

Within the ReConvert Editor, navigate to the widget you'd like to test

Click on the A/B Testing tab to enable A/B testing on the widget



A/B Testing can also be activated within the Flow Builder itself as can be seen in the following image.
If you already know which widgets you will offer and at what traffic percentages, then it may make more sense to create the A/B test here within the Flow Builder, set the traffic distribution as at 6. below, and then return to steps 4. and 5. (Either way will work and this is up to personal preference).



You will now see that a new menu has opened at the top of the preview panel. The % split is provided along with the name of the variant. You can click on the different variants and design exactly what you would like to test within each variant split. You can also add more variants to test by clicking add variant



Modify the variant with changes you'd like to test (e.g., title, design, offer - pretty much anything you would like!)

Allocate traffic percentages for each variant (e.g., 50% to Variant A, 50% to Variant B). Traffic allocation can be set within the Flow Builder (soon to be available directly in the editor)



To delete a variant you have created, simply expand the relevant variant within Flow Builder and click the trash button



Save the configuration and activate the test. The system will automatically show the appropriate variant to each user

Worked example

Here’s an example of a setup for an A/B test with four variants: Variant A, Variant B, Variant C, and Variant D.

If the customer meets the conditions set in the Flow Builder:

Variant A is shown to 41% of customers
Variant B is shown to 41% of customers
Variant C is shown to 8% of customers
Variant D is shown to 10% of customers

Variant examples:

Variant A: A product upsell of a green shirt, with the text in bold, medium sized and center-aligned



Variant B: A product upsell of a green shirt, with the text in bold, extra large text, left-aligned and with a hyperlink to the product page



Variant C: A product upsell widget - featuring one of our best selling products ('best sellers category') but excluding products from our 'New Product' category one after the other, keeping a consistent format to both in line with Variant A, with text indicating the nature of the promotion





💡 Tip: Understanding Product Categories
Best Sellers Category
This is an automated category created in the back-end based on the products most purchased by your customers.
New Product Category
This is a manual category we created. Learn how to create your own product categories by checking out our helpdesk article and video here.

Variant D: A product upsell widget with the same products as Variant C (best sellers, with new products excluded) but this time with greater emphasis on the text - the text is in bold, extra large text and left-aligned (similar to Variant B, but this time with a different product offering and text)

Monitoring and analyzing results

To determine the winning variant:

Within the Flow Builder, click the down arrow next to the variants in the right sidebar to see your results



Review performance metrics at the variant level such as:

Revenue
Impressions
Clicks
Conversions
CVR



Review performance metrics at the individual widget level within each variant, by clicking the down-arrow that appears next to widgets



Once you’ve gathered enough data, select the best-performing variant and apply it permanently to your widget. Decide whether to A/B test other variants to continually optimize your store's performance!

💡 Effective A/B Testing Tips
🔍 Test one element at a time: Focus on a single change (e.g., headline) to clearly understand its impact
⏳ Run tests for sufficient time: Allow enough time to gather meaningful data, especially if your store has low traffic
📊 Rely on data: Always let performance metrics guide your decisions—avoid using personal preferences. Let the data lead you to success

FAQs

Q: Can I test more than two variants?
A: Yes, you can create multiple variants.

Q: What happens to traffic distribution when I pause a variant?
A: The traffic will automatically redistribute among the remaining active variants (and the redistribution can be amended as per 6. above).

Q: How do I know when to stop a test?
A: Stop a test when you achieve statistical significance or when a clear winner emerges from the data.

By utilizing A/B testing, you can optimize your widgets and maximize their impact on your store’s performance.

If you have any further questions, feel free to reach out to our 24/7 customer care team by using the link in the bottom-right corner of any page within ReConvert.

Well, that’s a wrap!

If this article did not solve your issue, our support team is available for you 24/7, just click the live chat button at the bottom-right corner of this page.

📢 Did you know we have a YouTube channel? Don’t forget to subscribe to the channel to keep in touch with us and our news and updates.
🙏 If you find our app useful, help other merchants find it by leaving us an honest review on the Wix App Market! It means the world to us.
📧 Please let us know if you spot any typos, mistakes or omissions via care@reconvert.com

Updated on: 23/01/2025