A/B Testing - Why it's ok to be wrong.
So you don’t know if the headline to your website is attracting any traffic. You also might not know if your daily email has converted any online visitors, or if any of your marketing strategies are working, and if so, which one is working the best? Luckily for us, there is no time like the present where we have an abundance of data allowing us to make data driven decisions every day, and not having to worry about “guesstimating” in a business environment where so much is at stake.
Whether you are a digital marketer or not, A/B testing holds a lot of value and allows for constant improvements. A/B testing is a great way to determine what works best. Essentially, you are comparing two versions of the same webpage or app against each other to see which one is performing at a higher level. The two versions of a webpage are shown to users at random, and the differences can sometimes be very slight, like a different colored button on the webpage. While sometimes, the differences can be very different, such as a whole new webpage design. The end goal is to improve a specific website metric, such as clicks, bounce rate, or conversions.
This process allows individuals, teams, and companies to make changes to their products/services to improve their user experience, all while collecting data on the results. This allows us to see what we did wrong, or what we once thought was right, was actually not right at all. For example, if we wanted to use A/B testing to test the effectiveness of a google ad we created, we can learn which ad attracts most traffic and gets the most clicks (The original or the variation). If we wanted to learn which web layout has a higher conversion rate, then we would simply make a variation of the layout and test the subsequent landing page.
As users interact with both variations of your website, their engagement is measured and collected in an analytics dashboard, where the data is analyzed for us to determine whether changing certain aspects had a positive, negative, or no effect on user behavior.
So you want to conduct an A/B test? Here are the steps…
1. Identify a goal: Here is when you pick what you are looking to test. (ad, webpage, email)
2. Collect Data: This is when your analytics will provide insights into where to begin optimizing.
3. Hypothesis: it is important to write down what you think the outcome will be.
4. Create variations: Make the desired changes to certain elements to your website user experience. (Change the color of a button, customizing the page design, swapping web page elements, etc)
5. Run Experiment: Now users can visit both of your variations at random and their interactions with your website is measured and compared to the other.
6. Analyze Results: Once you have completed your experiment, you then analyze your results and derive deep meaningful insights.
After conducting an A/B test, and you find that your variation received better results than your original, then consider it a success, because there is always room for improvement! AB testing gives us room to be wrong, because were just a simple test away from turning that around.
Here are a few examples of companies who have successfully used A/B testing to improve their business.
1. HubSpot’s Lead Conversions: This company wanted to figure out what worked better a sign-up form inside a blog post or a separate sign up page? On one variation, the form was embedded in the post, while the other variation was a link to a landing page with the form. The form which was embedded in the post performed 71% better.
2. Google – Tests promo banner on mobile: Interstitials are the annoying full-screen ads that pop up on your website and cellphones. Supposedly, these annoying ads are meant to try and convert visitors into downloading their app, but it is rather the opposite. After conducting an A/B test, Google found that 69% of people left the website right away while only 9% of visitors clicked on the “Install” button. After Google implemented a less obtrusive app ad, active users increased by 17%.
If there is one thing you need to know, it is that good businesses and individuals always test their work. As just discussed, big companies take advantage of AB tests, understanding the value that comes along with being wrong. In order to succeed, you first must fail! I like to think the harder you fail, the harder you will succeed.
#wwudigimark