Getting the right kind of traffic to your website is essential for any ecommerce business. Learning how to build pages to attract those visitors is not an easy task for anyone, even the experts. A good SEO strategy can help improve visibility, but it can be tough to figure out how to improve your SEO.
One effective way to select an SEO strategy is to test out changes through an A/B test. It's easier than you think—all you need is your existing Google Search Console or Google Analytics and a bit of patience.
What is an A/B Test?
An A/B test involves presenting two different versions of something to consumers for a specific period of time. Comparing the results of pre-selected metrics allows you to determine which version is more effective and should be implemented moving forward. A/B tests are used across many types of actions, including email marketing and SEO.
Why Run an A/B Test?
When it comes to SEO, an A/B test allows you to compare your existing website strategy with the one you want to try out. It can help you to hone your strategy and fully roll out what works if you've seen a positive effect on a few pages.
Importantly, an A/B test prevents an SEO crisis. Imagine if you were to update SEO across the board, only to find out it had a negative effect on ranking and traffic. That is a much harder problem to fix. With an A/B test, you can identify the best new strategies and preserve what's already working.
How Do You Run an A/B Test?
Running an A/B test correctly involves a few important steps:
Step 1: Define the Parameters
Before you start your test, it is important to choose what metric to monitor so you know what to analyze to determine success or failure. While ranking may be important, so is the amount of traffic. Google Search Console and Google Analytics allow you to see both.
If your ranking stays the same, but more people are clicking through as the result of your change, the experiment may be a success. For this reason, try to look at both of these core metrics.
Step 2: Choose Test Pages
One thing you cannot do in an A/B test is post two versions of the same page at the same time—Google will see this as duplicate content and penalize both in the rankings. In addition, there are many factors that affect how Google sees a page. Two identical pages may not actually be identical from a technical standpoint—the original page may be older and exist in a different spot in your website hierarchy, so by creating a "copy" of the old page, you would not actually be testing the result of your SEO changes alone.
Instead, select a group of pages to test. For example, you may choose to put 20 product pages in your test group. Then, divide those pages into group A and group B. Group A will stay the same, acting as your control group. In group B, you will introduce changes. You can use a simple spreadsheet to track which URLs are in the test group and which are A or B.
Step 3: Select Changes to Test
You may have taken advice from an SEO expert or performed your own research into changes that can improve your rank and clickthrough rate. An A/B test is a good time to put those changes into practice.
On your spreadsheet, document the specific changes you will make to group B—those may include new title tags, meta descriptions, keyword density, and eliminating duplicate text. Implement the same changes on all group B pages so that there is no confusion about what factors cause what results. Your A/B test should give you an accurate picture of what will happen if you do an SEO refresh over your entire ecommerce site.
Step 4: Track Results
Before you run your test, predict the ranking and traffic for both your group A and B pages without any changes. You can use your past data on these pages to make these predictions. This way, you can identify whether a drop or boost in traffic is because of the experiment or because of other factors that affect all pages.
Then, run your test and compare your metrics on all pages to your predictions. Remember to look closely at both the control and your experimental groups to get a clear picture of the data. If pages in group B had a 10 percent increase in traffic and pages in group A had a 5 percent increase, you might reasonably assume that the difference of 5 percent was because of your SEO changes.
Step 5: Note What's Surprising
Sometimes, following SEO best practices has surprising results—a change meant to improve ranking could hurt it, or pages may rank the same but have fewer hits. By taking a close look at your data and identifying what is surprising, you can continue to tweak the strategy to find what works.
Running an A/B test is an effective way to improve your website's SEO. You don’t even need any special online tools to perform an A/B test—just the tools you have already been using to monitor your site's traffic and ranking. Done in a controlled manner, you won't have to worry about losing the traction your site already has as you decide your future course of action.