Whether you want to A/B test your opt-in forms, lead generation pages, sales pages, landing pages, blog post headlines or online quiz pages, it’s important to apply the 80/20 rule to your planning and design.
That means taking a moment to prioritize your most important test variables that offer the right balance of conversion optimization leverage for the least amount of time spent crafting A/B test variations.
For example, testing two different button colors for an opt-in form or landing page takes seconds to do, but will have no real impact on your conversion rates. Worse, it won’t teach you anything meaningful about your audience.
On the other hand, A/B testing your pricing on a sales page is a variable you can quickly change with a huge potential conversion and revenue impact.
See the Time To Create Variations vs. Conversion Rate Optimization (CRO) Lever Size matrix in the next section to get an idea of how to prioritize your testing variables to get the most impact from your A/B tests right away.
Create Your Own "Time To Create Variations vs. CRO Lever Size" Matrix
The matrix shown below plots several different A/B test variables based on:
- An X-scale judging how long it takes to create variations for a particular variable and,
- A Y-scale judging the conversion rate potential a particular variable offers.
Testing an element like an image slider will take lots of time and effort to set up with very little CRO upside. Testing a different headline however, takes very little time to do and has the potential to increase conversions significantly.
Study the matrix below to get an idea what test elements you should prioritize first and which variables to avoid wasting your time with:
Take a few moments to brainstorm your own matrix to add any additional test elements or variables you're interested in A/B testing.
If you'd like to see some additional A/B testing ideas, download our free A/B Testing Examples Reference PDF by clicking the button below:
Click the next lesson arrow below to move on to Step 2 in the toolkit.