Why You Should A/B Test Your Marketing Campaigns

February 19, 2018 Natalie Russo

Companies spend $92 on average to convince a potential customer to visit their website, but only $1 so that the website converts that potential customer into a real one. Doesn’t make sense, does it?

One explanation is that business owners are unaware of what to change on a page to obtain better results. This is where A/B testing comes in handy.


  • Number of people using Google to do a keyword search: 10,000
  • Number of people clicking your link and visiting your website: 1,000
  • Number of people making a purchase: 20

In this example, only 2% of visitors become customers. In other words, you convert 2% of your traffic. It’s not a bad score, but if it increased to 3%, you would get 10 more sales and increase your revenues by 50%!

Would it be more effective to replace the photo of your team with a photo of your product? Would it be better to change the colour of the button that web surfers need to click? A/B testing gives you a clear and simple answer.

If you’d like to test the colour of your button, there’s software that will make sure that 50% of visitors see your current button (let’s say it’s green) and 50% see your new button (let’s say it’s red). You let the test run for as long as it takes to collect enough data, then after a few days or weeks, you compare the statistics for the two groups.

  • Group A (green button): 2.00%
  • Group B (red button): 2.15%

Now you have your answer: the red button is more effective! So you end the testing and you use the red button permanently.

You can then start a new test where you alternate the photos.

• Group A (team photo): 2.15%

• Group B (product photo): 1.90%

So you keep the team photo.

By continuously conducting this kind of testing, you will improve the effectiveness of your site, and therefore your profits.

Where do you start?

Google, as has become typical, offers an A/B testing tool free to everyone: the Experiments section of Google Analytics. You can also buy software that may be easier to use or more powerful.

Elements to test

  • Page title
  • Words used on your button
  • Button colour
  • Button position on the page
  • Length of text on the page
  • Number of questions in your form
  • Whether or not to post prices
  • Whether or not to offer discounts
  • Which photos to use
  • Whether to add a video

In the example above, keeping the green button and using a photo of the product lowers sales by 10%, while changing just the button increases sales by 15%! This is a perfect example of why you should only test one thing at a time. You can test several elements on the same page, but interpreting the A/B statistics becomes much more complicated.

You can also conduct several A/B tests at the same time on different pages, but make sure that the tests do not interfere with one another.

How many visitors should you test? That depends on many things, and good A/B testing software will give you clear instructions to use for your situation. As a general rule, the less traffic your page receives, the longer you should run the test.

Start with a very simple test, like a button colour. Soon, A/B testing will be like a game. Some companies ask their employees to predict the results and give prizes to those who predict correctly. What might at first seem daunting quickly becomes routine, and before you know it, you will always have A/B tests running, and you will gradually see the revenues generated from your website increase.

Need advice on this topic to see how it can help your business thrive? Click here to talk to an expert.

Previous Article
6 Biggest SEO Changes to Know About in 2016
6 Biggest SEO Changes to Know About in 2016

Search engine optimization, known more commonly as SEO, is a lot more than selecting the right keywords. As...

Next Article
5 Big Website Mistakes to Avoid
5 Big Website Mistakes to Avoid

Your website is the go to place when customers want to learn more about your business, so it is especially ...