Conversion Optimization: A/B Split Testing 101 for Beginners
Take a moment to imagine the following hypothetical situation…
You own a business that sells furniture online. As the spring approaches, you decide to run a spring clearance sale and advertise it on your website’s homepage.
You already have all your ducks in a row in terms of the copy and layout of the announcement on your site, but you have yet to decide on a Call to Action (CTA)–a small yet significant factor that can affect click-throughs and conversions.
You’ve created two possible options for CTAs–but how do you choose between them with confidence, knowing the one you select will drive the most conversions?
Enter A/B split testing.
What A/B split is and how it works.
A/B split testing is an effective and common method of comparing two versions of a digital element against each other to determine which one performs better.
The method can test emails, webpages, apps, and other digital properties to optimize conversions and generate increased revenue.
From Amazon and Google to small- and medium-sized businesses, all sorts of companies deploy this method of testing to evaluate and guide their marketing decisions.
Here’s how it works, using the furniture company example:
To determine which CTA to use, you would run a test in which half of the visitors would see CTA A (the control) and the other half would see CTA B (the variation).
Visitors who would be shown CTA B would unknowingly be redirected to a different landing page during the period of the test.
The CTA that brings you the most clicks and subsequent conversions would be the most effective one to use on your homepage for the remainder of the marketing campaign.
Why A/B split testing is helpful.
The fundamental value of an A/B test is that it helps business owners and marketers optimize for conversions. Here are a few additional reasons it’s valuable…
You can make more out of your existing traffic. Acquiring paid traffic can be costly. On the other hand, increasing conversions with existing traffic using A/B testing involves minimal costs.
It can encompass a number of variable elements. The example we gave tested only one variable (the CTA), but split testing can be applied to an entire landing page, testing various elements at once–such as headlines, paragraph text, testimonials, and images.
It determines how people actually act. Focus groups and user testing ask users how they think they would act when served two different options, which leaves room for bias and error. Split testing puts tests subjects in situations where they aren’t aware of the test and act as they normally would.
It can save you time. The amount of time between launching a webpage and generating revenue from it can be lessened with split testing. Rather than launching a landing page and waiting to see how it performs, split testing allows you to quickly identify elements that work and don’t work when optimizing for conversions.
It enables you to make data-informed decisions. Split testing takes the guesswork out of website optimization and allows you to feel confident in making decisions for your business.
Split testing tips to keep in mind.
You may be excited to start split testing your webpages and emails, but before you do, here are a few important pearls of advice you’ll need to bear in mind.
Track and monitor performance. Be sure to track variables such as opens, clicks, unsubscribes, and conversions so you’ll have the data to support your decisions after the testing is over.
Remember seasonality. Avoid running tests during seasonal periods where the results may be affected by natural changes in consumer behavior.
Run the test long enough in order to get enough sample views. According to HubSpot, at least 100 views each usually makes for an accurate test, especially if you’re only testing one variable on a page (such as a CTA).
Split testing can harm SEO if done improperly. When split testing a webpage, you need to ensure the variant page in an A/B test cannot be indexed by Google to avoid any negative consequences (such as the duplicate content penalty).
In the hypothetical situation we described earlier, the split test determined that CTA A drove 25% more conversions than CTA B.
If you want to improve your real-life marketing dollars’ ROI, we suggest you begin applying this method to your business as soon as possible.
If you’re thinking of split testing your emails,
click here to read about certain email elements that can lead to higher opens and lead to conversions.