A/B testing is a marketing strategy which involves comparing two versions of a page to determine which performs better in test markets. A/B tests can be performed on platforms such as web pages, emails, or mobile apps. While A/B testing has been around for awhile, it is becoming a more common tool in the digital age because the costs of are low, and the data collected can provide useful insight to consumer behavior. If you wanted to perform an A/B test on a web page, you would modify the current version of that page to create an all new second version. The change could be something small such as a different headline, or it could be a complete website overhaul. The current version of a page serves as the control variable, allowing you to easily compare the data collected between variants of pages. Once you have your variants established, your traffic is randomly disbursed to the different versions of your page. Data is collected in an analytics dashboard, and you can use it to determine which version of a page had more success. This data can help reveal what is needed to be done to optimize a marketing campaign.

When performing an A/B test, there is a helpful outline of steps you can follow so you get the most out of the experience.

1. Collect Data: Using analytics software such as Google Analytics, you can begin collecting data to determine potential areas of improvement. If a particular page on your website is generating a low number of conversions/clicks compared to other pages, that may be a good page to run an A/B test on.
2. Identify Goals: You must identify goals to determine which version of your page is more successful. Your goals could be something like increasing sales on a particular page, or even increasing the number of times users click a specific link.
3. Generate Hypothesis: Generate a list of hypothesis that could lead to you reaching your goal. For example, if your goal is to sell more running shoes on an ecommerce site, you may predict that adding an inspirational running photo to the page will increase sales.
4. Create Variations: Based on your hypothesis, generate variations of your current page that will be used in test markets. To do this, use your A/B testing software of choice and make the desired changes to your page.
5. Run Experiment: Launch the control page along with the variants to start collecting data! Visitors will be assigned to a random page, and data will be collected to help analyze their experiences.
6. Analyze Results: Upon completion of your experiment, data will be presented to you on how the various versions of your page performed. You can use this data to infer if the variation was a success, or you may decide you want to test new variants of the page. In this case, you would repeat the process to collect more data.

Many successful businesses have taken advantage of A/B testing. Paperstone is a UK based ecommerce website that offers office supplies. After an A/B test, they discovered that displaying competitor’s higher prices on product pages increased conversion rates by 10%. 160 Driving Academy is an Illinois based truck driving school. They discovered that replacing a stock picture with a picture of an actual student on the homepage increased conversions by 161%! Media giant Electronic Arts has also taken advantage of A/B testing. EA wanted to maximize revenue earned from the release of its video game titled SimCity 5. This case study was especially interesting because EA discovered that taking away a pre order incentive ($20 off a future purchase) actually resulted in a 40% sales increase when compared to the control page. It turns out that fans of the game didn’t want the page to include any incentive, they just wanted to purchase the game as easily as possible. A/B testing can help us gain insight into customer’s minds. It is a powerful tool that should surely be taken advantage of to optimize your business.