- Magnetic Marketing
- The right way to A/B test
The right way to A/B test
If you ain't testing, you ain't trying.
Welcome to the 29th issue of Always Be Learning. The newsletter that holds the door open for you, even when you're far enough away that it creates a few seconds of awkwardness.
If you were forwarded this email, you can subscribe by clicking below:
Friend of the program:
Alex & Books Newsletter - Get a new book summary every week with actionable advice to improve your life.
Join 24,000+ people who love reading by clicking below!
Today's Marketing Breakdown:
In this week's issue, I'm going to show you how to run a proper A/B test.
I see marketers mess this up all the time. And the results are painful.
It leads to:
Wasted time and resources
The Big Mistake:
Trying to do too much at once.
I get it. You want to move fast. You want to learn quickly. But you can't rush this process.
Most small businesses only have the resources to test one variable at a time.
Why is this important?
Because if you're changing multiple variables at once in an A/B test, you'll have no idea which variable actually made the difference and drove the uplift.
How To Identify Winners:
When looking for A/B test winners, the main thing I'm looking for is statistical significance.
That is usually represented with a 90% confidence score or higher (100% is best).
I won't get into the mathematics behind calculating this metric because I'm no math professor...
But in basic terms, I'm looking for:
A big enough sample size
A big enough difference in the key metric being measured
If you're interested, feel free to steal my A/B test calculator that I built in google sheets to calculate statistical significance on my email marketing A/B tests (use this for Open Rate and Click To Open Rate tests).
There are instructions on how to use this calculator in the second tab, but if you have any questions, just shoot me a reply to this email!
There is a world where you can test multiple variables at once - this is called Multivariate testing.
Let's revisit our hypothetical testing situation from earlier in this newsletter: You want to run a test on your homepage creative by testing the image AND the headline.
As stated above, the simplest way to do this would be to run one test at a time.
However, if you have the resources, you could test both at the same time. To do this, you just need to set up 4 variations (instead of 2) to account for all possible combinations (see below for an example)
With this setup, you're knocking out 2 tests at once. You just need to wait longer to collect enough data to reach a significant sample size.
As you can imagine, manually running these multivariate tests could be time-consuming and leave plenty of room for error.
Luckily, some tools out there can help facilitate these tests with less effort on your part.
Some tools to check out for multivariate testing:
A/B testing is your best friend - when used correctly.
For most, that means testing one variable at a time to keep it simple. These small wins add up and compound over time.
Feeling advanced? Give multi-variate testing a shot when you have the team and resources ready to execute.
There you have it! 1 marketing tip to help grow your DTC business.
Here’s what else I’m checking out in the DTC marketing world:
Hope this has been helpful!
If you enjoyed it, please forward it along to your marketing friends.
See ya next Thursday,
P.S. This is the final call for my Email Marketing 101 course!
Cohort 2 of my Email Marketing 101 course starts this upcoming Monday (9/26).
I'll show you how email marketing can help your business:
Convert more leads
Lower customer acquisition cost
Increase customer lifetime value
Retain & upsell existing customers
What business can afford to ignore that?
Grab your seat today! 👇
By the way, everyone gets a 100% money-back guarantee. If you take the course and don't get value, I'll give you a full refund.
Hope to see ya in the course!