Let’s put your brand to the test. A/B testing that is.
Did you know that Google tested 41 shades of blue? Although small, these seemingly minimal choices can help your business boom, but these choices need to be based on one thing, data.
But you don’t have to be a giant company to implement testing. Read on for an easy how to in A/B testing 🙂
So here’s the simple explanation:
The goal is to make small improvements to your website, app, and even emails. You basically create variations of what you want to test and you send it out into the world.
Each test will give you the information you need to better understand your clients and optimize to their needs.
In order to perform a simple test, you’ll need access to at least a few basic analytics. These include click rates, open rates, and the ability to create slightly different versions of the item you’re testing. See below for a list of resources.
Let’s say you’ll be testing an existing email.
Pick a portion of the email that you need to improve by checking your current stats. Do your analytics show that your clients aren’t clicking on the offer? Choose something that is not resonating with those receiving your email.
The secret sauce:
Only change one thing at a time. You might think you can completely revamp an email or change just 3 little things. Don’t do this. It can potentially mess with your analytics.
Again, keep it simple and choose just one thing to edit. For example: let’s test a call to action on your email. This doesn’t mean change the color, copy, and position. Perhaps only test the copy. Instead of “Shop Now” try “Shop”, “Get Sexy”, or even “Get your shit now!”
Now it’s time to choose how many users you’ll be sending out your test to. Some tools simplify the process and don’t allow you to choose who gets the test or you could do that yourself and choose how long to run it for.
Variables for choosing the length of time and who will see the a/b test depends on the traffic your typical email gets and if you’d prefer to keep it on the down-low. Check out the resources below for choosing the length of time.
Ask yourself how large a test has to be for you to truly place your trust in the findings. The ultimate goal is to tell yourself that with at least 95% certainty that the change you made helped make a choice in the color/layout/copy.
Now send it out! I know, it sounds too simple and silly but this is how you’ll slowly start optimizing for your clients.
But don’t sit back! The work isn’t over yet. Grab your coffee and start looking at your analytics. Warning: don’t get DISCOURAGED.
It’s normal to see very little change in your test. Remember that this type of testing is to learn about the small changes you can make to better connect with a user. It’s also PERFECT when you see a negative change.
No, no, I haven’t gone crazy. You see, getting a negative change will let you know what you shouldn’t do.
Okay, so you’ve gotten your confirmation and now you have a choice: you could officially roll out your change or continue testing a different version of your email.
There’s a few things to keep in mind when doing a/b testing. It won’t tell you why things worked or didn’t work. Testing won’t work if you’re doing it with large changes. And, finally, it won’t replace in-depth user research.
A/B testing isn’t just for the big shot companies out there. It’s perfect and important for small businesses to survive.
Resources for Data:
Resources for checking how long to run your test: