In this video post, Hanne debunks 3 MASSIVE myths around conversion rate optimization and A/B testing.
Read on or watch the video to learn what they are and why they’re not true...
Myth #1. You Need a Degree in Statistics to Do CRO
Not true at all!
We now live in a world where your tools and software can handle all the advanced statistical calculations for you so don’t let intimidating maths scare you away from doing Conversion Rate Optimization (CRO) — like A/B testing — on your WordPress website.
The only thing you need to do is understand how to use your CRO tools and understand how to interpret the results they give you.
To do that, it’s important to know 2 important concepts...
Concept #1: Randomness
How do you know if the results of any A/B test you run on your website reflect something real about the designs you’re testing instead of something totally random?
In other words, how do you know the conversions (as well as non-conversions) you get occur due to the actual design changes you made and not because your visitor would have signed up (or not) regardless?
When the reason a visitor converts is not related to the changes you made between your A/B test variations, we call that factor randomness. In CRO, it’s important you have a way to make sure your results aren’t random, and that’s where the next concept comes into play...
Concept #2: Statistical Significance
How can you feel confident that the results of your CRO experiments and A/B tests aren’t random?
You need to achieve a certain confidence threshold known as statistical significance.
Without getting into the weeds about how it’s calculated, most A/B testing tools will display statistical significance in terms of a unit-less percentage. And once a certain percentage threshold is achieved by your test, you can confidently select a winner.
Here at Thrive Themes, we suggest that your A/B tests (whether they're landing page A/B tests in Thrive Optimize or opt-in form A/B tests in Thrive Leads) achieve a statistical significance of 95% or higher before trusting that the your CRO results are real and not random.
Myth #2: You Need a Lot of Traffic to Do Conversion Rate Optimization
Although higher traffic volumes do help generate more data faster, it’s definitely not required for creating conversion boosting A/B tests.
In fact, the key to creating A/B tests for low traffic websites is designing two VERY DIFFERENT test variations.
To illustrate this idea, think about running a simple experiment where you ask people “Which is heavier… A or B?”
If A is an apple and B is an orange, you’ll need a TON of data to reach any sort of statistically significant result (because the weight difference between most apples and oranges is so similar).
But if A is a paper clip and B is a bowling ball, the weight difference between these two objects is so obvious to most people, you’ll reach a statistically significant result after asking a lot less people.
Learn more about effective A/B testing design on low traffic websites by reading this blog post.
Myth #3: You Need Complicated & Expensive Tools
Want to see how easy it is to debunk this myth?
Just check out how affordable Thrive Suite is, which includes A/B testing tools throughout the toolkit, including:
Use Thrive Optimize to A/B test your landing page designs to boost conversions on your sales and lead generation funnels.
If you’re not a Thrive Suite user, you can also use a free A/B testing tool like Google Optimize to get started running CRO experiments.
Google Optimize is definitely more complicated to set up, but the documentation is good enough to help you get set up without too much hassle.
Time to Start Testing
No matter where you are in your website building and online business journey, now you know how easy it is to get started running conversion boosting A/B tests and CRO experiments on your own website.
And if you’re keen to see just how easy it is to set up landing page A/B tests on your WordPress site, check out this short walkthrough video showing how to use Thrive Optimize:
Have you stumbled across any other CRO myths that we didn’t discuss in this video?
Share them with us in the comments below!
Hi Hanne, good points you made there although I do find that some of my tests seem to run forever and never complete.
I guess this is because certain pages/forms I want to test must fall below a minimum traffic level required to get the 95% statistical significance – in a reasonable time period.
How much traffic would you say is the minimum for a decent A/B test along the lines of the feather/bowling ball analogy?
Let’s say a 100 visitors per variation but again that depends so much on what you’re testing 🙂
I’m assuming that the CRO rate is at least above 1% and hopefully much more.
Eg. We’re running tests on landing pages that get 40% signup rates. With a 100 people to each page you can see the difference (or you can see that they’re performing very similarly and will not get to a good conclusion and the difference should be bigger).