A/B Testing is where you create two versions of the same webpage, social post, email, etc. The idea is that you can then run them for a pre-defined period to see which one visitors/users respond to better.
In simple terms, A/B testing is where you put two sunflower seeds* in their own plant pot and give them different growing conditions** to see which one did better.
*aka webpages
**different content, different imagery, different CTAs, etc
The key with A/B testing is to ensure you have all of the analytics and tracking in place so you can record all interactions with each version to accurately see which version performed better.
Stats that you might be looking at in A/B testing include…
- Volume of traffic on the webpage/social post/email
- Bounce rate of the webpage/social post/email
- Click through rate of links on the webpage/social post/email
- Measure of any forms completed
- How many purchases were made
A/B testing allows you to try things out – see what works and what doesn’t work – allowing you to fine-tune your website for maximum conversions.
Frequently Asked Questions
What do I need for A/B testing?
For A/B testing you need…
- A webpage, social post, or email to test
- Anywhere between 1 to 3 variations within that webpage, social post, or email that you can change to prompt a better conversion rate
- Defined goals for that A/B testing - knowing what metrics you are going to track and having that tracking in place
What elements could I change?
You can change anything on your webpage/social post/email to see whether it has a positive effect on your conversion rates. They include…
- Content
- Layout of Content
- Use of headers/subheadings
- Use of Imagery/Video
- Colours/Contrast
- Number of Images/Videos
- Pricing / Highlighting of special offers
- The CTAs used and their design
Can A/B testing only take place with two versions?
You can put more than 2 versions of a webpage/social post/email out there under the principles of A/B testing and still refer to it as A/B testing.
However, you need to balance the volume of variations out with how you can tell what one did better. Too many variations mean that you may not be able to pinpoint exactly why one performed better than another.
This is why it may be better to do two versions, review them, then take the winner and do a further two versions. Then review, and then take the winner of that round to make another two versions, and so on. More control equals better data and that means your business can make intentional and controlled steps forward in a positive manner.
Is A/B testing a new thing?
A/B testing is not a new thing, it has been around since the 1960s. Sometimes referred to as ‘Splitting’ or ‘Bucketing’.
If you are looking to improve your website, social posts or emails then get in touch with Digital Nachos today. We can review your current offering and work with you to see improvements.
Published: 5th Feb 2024