Exploring A/B Testing

hand-457335_640One of the biggest questions small business owners have about their online marketing is, “could we be doing this better?”

You want your website, landing pages, sales pages, and online ads to be as effective as possible. You might find yourself questioning the images, copy, layout, colors, fonts, buttons, the amount of text used, pricing strategies, and other elements on each page, wondering if more people would opt-in to your offer or buy your product if something about the page was different.

“Maybe the headline didn’t grab their attention….”

“Perhaps the photo we chose didn’t convey the right feeling…”

“Would a yellow button get more clicks?”

“Should we include the price?”

It’s common to second-guess your choices and wonder if a little tweak ‘here’ and a slight change ‘there’ might result in higher conversions. After all, your ultimate goal  is to ensure your marketing efforts are successful and profitable.

So how can you gain more clarity about what is performing well and what can be improved?

A/B testing, also known as split testing, is a way to get the best performance out of your web, landing, and sales pages. Put simply, A/B testing, starts with two test versions of a page – version A and version B – each with the same goal, which serves as a metric to measure success. Version A serves as your ‘control’ in the test, which might be your existing page that you are hoping to improve. Page B serves as the experimental version wherein you test variables.

For example, both site A and site B might invite your site’s visitors to sign up for a free consultation. To test which version gets better results — in this case, more requests for a free consultation — you would send some of your visitors to site A and some visitors to site B.

Assuming site A is your existing page, site B would include one or, at the most, two changes. You might add an arrow pointing to the submission form, change the term ‘consultation’ to ‘discovery session,’ add a photograph of yourself next to the form, or add a bright-colored submission button.

The next step is to run two versions simultaneously and measure the number of requests per page. Here are some examples of what could happen next:

If version A uses the term ‘consultation’ and version B uses the term ‘discovery session,’ and you notice fewer people opt for a ‘discovery session,’ stick with ‘consultation’ – that is not a change you should make.

If version A does not include your photograph and version B includes a friendly professional headshot, and you find that the headshot prompts more requests, you know your smiling face is a keeper. Now add your photo to the new version A.

Now that you are using your photograph in version A, version B will also include your photo. But now, try changing the color of the submission button from gray to a vibrant green. If the green button gets more clicks, keep the green button and add it to version A.

Now version A has evolved to include both your smiling face and a bright green button. Why not add a big green arrow to the experiment as well? Maybe that will capture the attention of more visitors. Run your A/B test with the green arrow on version B. Did more visitors sign up with version A — the one without a green arrow? The green arrow must have been too much — perhaps it cluttered the design or it turned off your visitors in some way. Ditch the arrow, and keep going.

This is just one simple example of how you can use A/B testing to optimize the performance of your website. You can use a similar process to test your bounce rate (time spent on a page), click throughs, sales, and more. You can also use split testing on advertisements, social media, and nearly every aspect of your online marketing.

If you want to use A/B testing to increase the success and profitability of your marketing efforts, be sure to keep the following tips in mind:

  • Always test two versions simultaneously. If you run version A one week, and version B the next week, you have already changed a variable (timing), which can muddle your results.
  • Keep your variables to a minimum. Only change one element at a time, or two at most. Otherwise, you are not going to know which tweak contributed to different results.
  • Let the test run long enough to gather sufficient data. Sometimes you might feel anxious about exposure while testing an element, but be sure not to abandon the experiment without giving it a chance to run its course.
  • Start with a plan. In advance, know exactly why you are split testing, what you are split testing, and what results you hope to gain.
  • If your goal is to optimize for conversions, let the testing serve as the ultimate guideline. Just because you like the way version A looks better doesn’t mean you should ignore the fact that version B converts better.
  • Try not to let it get out of hand. Sometimes you can go overboard with split testing everything, to the point that you are compromising your big picture goals. A/B testing is not a magic solution for everything, but just one tool.
  • Be careful. Sometimes A/B testing can go awry if not handled properly. You need to be in control of who sees what and when. Otherwise, visitors can see conflicting or confusing information.
  • Get help. Doing A/B testing effectively is going to take an investment of time and possibly money. Get familiar with the process and technologies or hire a team like Solamar to run the testing for you.

With the right approach, A/B testing can ensure that your marketing and promotions efforts are getting the best results possible. Small businesses often do not have the budget for A/B testing technology used by bigger companies with large marketing budgets. We suggest that you start by making smart decisions about how to plan and run your A/B testing effectively. Contact us to get started.

Want blog updates sent to you like magic?

Just enter your email below.

Leave a Reply

Your email address will not be published. Required fields are marked *