As you explore the world of Inbound Marketing, you'll start to see more and more about the importance of testing. These concepts are critical to your Inbound success, but sometimes the best practices for implementation are a little hazy for traditional marketers and small business owners. That's why today we're going to discuss just how you can go about taking on the world of split testing, and start maximizing your brand's online potential.
First, I'll have to admit that I had a little bit of a hard time selecting an image for today's post. If you look around online, there are tons of articles written on A/B testing tips and best practices - and they all have pretty boring pictures.
I found the horses under the search term "two of a kind", in case you're wondering how those equine beauties are related to inbound marketing.
Alright, glad that's covered. Let's get down to business.
So, what is A/B testing, and what in the world is CRO?
It all goes back to conversions. As you know, a conversion is your visitor's participation in an intended action. Conversions could be click throughs, purchases, social likes, etc. – whatever it is you want that visitor to do.
Conversion rate refers to how many visitors actually partake in that activity. If you have 100 visitors per day and 20 convert, your conversion rate is 20%. A/B testing is a Conversion Rate Optimization practice which allows you to determine which activity on your part leads to the highest conversion rates.
A/B testing can be practiced in nearly all areas of your digital promotion including things such as:
- Call to action copy, sizing, or appearance
- Email subject lines, call to action usage, time sent, copy, images used, etc.
- Website layout
- Nature of promotional offers
All you'll need to do is determine what you want to test, create two (or three, or more) versions of it, and track which one performs better throughout a set period of time. Remember that performing, or converting, is based on your ultimate goal.
After partaking in a successful A/B test, you can optimize your material to deliver the highest conversion rate. And let's be honest – why wouldn't you want to provide an experience that prospects and leads have been proven to respond to?
This question explains why A/B testing is so popular among the most successful digital companies and marketers. But A/B testing is a science, and every experiment requires a steady methodology. So, let's take a look at a few things you absolutely must know in order to get the most out of your testing process:
- Test clean
Changing multiple aspects of your content is a major mistake to avoid in A/B testing. When you determine what you want to measure (color of button, email subject, etc.) you need to focus on that one variable in order to get legitimate results. Here's why:
Let's say you send out two emails, each with a different subject line and unique call to action in the body. Email 1 gets a 90% higher conversion rate – good for you! But how do you know what contributed to that increase? It's impossible to say when you have multiple variables, so keep your testing clean by only altering one aspect of each test
There is such a thing as multivariate testing, but it's a different process entirely. To learn the difference and a little more about why it matters, check out this article on HubSpot.
- Start small
Although it may be tempting to utilize A/B testing to completely revolutionize your online presence, it's a good idea to start by dipping your toe in the water.
Your first few tests will be the most productive if you start with simple aspects which are easy for you to measure. A great example would be to simply change the copy in a call to action, and monitor which one gets more clicks. Even that change can be minute: something like "Version A: Click here to download" VS "Version B: Download your guide here" can provide you with valuable insight.
As you become more comfortable with testing and the best ways to create different versions of your content, you'll be able to start trying multiple tests at a time and really digging into the different facets of your audience's behavior. But get familiar with the basics of the game before you tackle the big leagues.
- Don't make assumptions
A/B testing often reveals information about our audience which we have a hard time believing. Sometimes it's the unnatural sounding copy or less aesthetically intriguing picture which yields a greater conversion rate. Embrace it!
A common mistake among A/B beginners is to assume the testing is wrong, and our intuition is superior to data. Avoid this behavior like the plague.
Even when you're astounded by what you find, remember that numbers don't lie. It's much more valuable to deliver an experience which customers have been numerically proven to enjoy over one which you assume they will based on your own bias.
Stay true to the process, optimize based on data rather than personal perception, and you'll have conquered a major hurdle in the testing process.
- Stick with it
Sometimes, it takes a while to see clear or correct results from A/B testing. When you set up a test, you need to set a length of time for its duration – and stick to it. You'll be surprised how much results can change between hours or days.
One version may dominate the other one for a week, only to be suddenly and inexplicably overthrown a few days later. Always bear in mind that it ain't over til it's over, and stick to the schedule in order to extract the most precise and valid results at the true end of the test.
Now that you're familiar with A/B testing and CRO, I'm going to throw one more important acronym at you.
ABT refers to the fact that we as marketers should "Always Be Testing".
Although it may seem lucrative to run a test, optimize, and move on, it's incredibly important that you think of testing as a continuous process. The more you test, the more opportunities you have to uncover hidden insights about your website, audience, and content.
I love this quote from Search Engine Watch which summarizes what separates A/B testers from A/B "badasses":
These tips are a great foundation for you're A/B knowledge, but there are a ton of other resources out there which can help you along your way to ninja status. Here are a few of my favorites:
What will you measure in your first A/B test? Let me know how it goes in the comments below!
Image credit: http://www.flickr.com/photos/19901746@N00/5315857965