Most engineers and scientists swear by testing, testing, and more testing. And marketers are the same. In fact, Steelhouse proved that “using correct targeting and testing methods can increase conversion rates up to 300 percent.”
A/B testing is one of the best – and most used - techniques to learn what content performs best with your audience, but it can be hard to know how to start the process and choose the right elements.
With A/B testing, you create two or more versions of one marketing element, such as a subject line of an email or CTA button, then put it in front of two randomly selected audiences to see which performs better. The best part of A/B testing is that you are testing two segments of your audience, not relying on industry standards. Your contacts are the judges, and the outcome can significantly affect your conversion rates and, ultimately, your bottom line.
The Importance of A/B Testing
Marketers still struggle to prove the ROI of digital marketing, inbound marketing and content marketing. A/B testing is the simplest and easiest way to find out whether marketing efforts work or not.
There are many variables involved in the success inbound marketing and content marketing (such as SEO, quality of content, user experience, timing, social media, etc.) and these variables involve many elements and methodologies (such as link building, heading text, and images - just for SEO). So while it may seem ridiculous to simply test the use of an image in one blog post, it can reveal a lot about your audience and help you improve your marketing methods and tactics. One step at a time!
Getting Started with A/B Testing
Marketers are often intimidated by the idea of A/B testing, where you pit one type of content against another in a head-to-head battle to see which is more successful. It seems like double the amount of work. Additionally, it feels like there is a lot to test and not a lot of time or resources to test it. Just thinking in numbers, in general, conversion rates range from about 1-3%. This leaves an average of 98% of your traffic not converting – and therefore needing tested for what will improve that. But there isn’t one test to figure out how to push more of that 98% across. Think in incremental numbers. If you can get your conversion rate up to 4%, you’ve made major progress.
Here are some tips for getting started with A/B testing:
Tip 1: Pick One Variable at a Time
Don’t try to A/B test multiple variables at once. If you are trying to figure out a way to increase your open-rates for your emails, only use the subject line as the A/B variables, rather than using different subject lines AND sending at different times AND sending from different email addresses.
Here are some examples of simple variables to test and the metrics they reveal:
To Test the Metrics of Click-Through-Rates in Emails or on Landing Pages:
- Number of CTAs in an email or landing page
- Number of email columns
- CTA verbiage
To Test Email Open Rates:
- Subject Lines
- Sender name or email address (or both)
- Send time
- Send day
To Test Conversion Rates in Forms:
- Number of form fields
- Type of form fields
To Test Time on Site:
- Format of text (bullets, text, etc.)
Tip 2: Narrow Down Your Audience, but Don’t Limit Your Results
Whose opinion do you need? You may want to see how a group of contacts in a specific industry respond to one product launch email CTA vs. another, or maybe you want to see what resonates more with your C-level contacts. Try to have at least a few hundred contacts in your test group to make sure you’re getting statistically significant results.
Tip 3: Split the List and Send
Splitting your group down the middle into two equal groups is the easiest way to go. Remember, don’t sway the list one way or the other. A/B testing is like surveying; you want the two audiences to be similar in their variation of personalities. You wouldn’t throw a dog bone to a group of dogs and catnip to a group of cats and call that A/B testing on what treat animals like best.
**Pro-tip: For something like a product launch email, where it would be nice to know the best-performing content before the launch date, consider sending a test to 10% of your list a day or two before launch, then sending the winning content to the remaining 90% on launch day.
Tip 4: Analyze and Apply
The waiting is, as they say, the hardest part. After you send your test, don’t rush to figure out the results. Give it a few days (the exact time will vary) before comparing metrics and picking a winner. Now you have empirical evidence showing that one variable is better than another, and you can use this info when crafting future content, designing emails and landing pages, and so on. You’re practically a scientist now.
Here's an infographic all about the basics of A/B testing:
Need help planning your marketing automation workflows? Learn tips, tricks and best-practices for developing workflows.
This blog post has been updated from it's original publications on April 23, 2015.