How A/B Testing Can Improve the ROI of Your Marketing Campaigns

Written by on September 19, 2018

We all want to improve the ROI of our marketing campaigns.

You’ve worked hard on them. You and your team have created blog content, PPC ads, emails, landing pages, and more with the best of intentions. You probably know that cheap content can’t drive ROI. But the ROI you wanted to see just isn’t there.

Chances are, you’ve probably heard the saying, “The definition of insanity is doing the same thing over and over again, but expecting different results.”  The same is true for your marketing campaigns, your website, and your content.

To boost ROI and get optimal conversions, you’ve got to test different variables and find out what your audience actually responds to.

And A/B testing is the way to do it.

What is A/B testing?

An A/B test consists of creating two or more versions of a piece of content like a landing page, email, or ad and modifying one element in each version to see which one performs the best.

And the potential ROI is huge. Yuppiechef increased conversions on one of their landing pages by a whopping 100% by merely removing a navigation bar in an A/B test.

It’s even possible to reach an ROI of 567% or more by increasing conversions with A/B testing — if you play your cards right.

Elements to test include colors, images, buttons, text, the call to action, or anything else that has the potential to impact user experience. When designing these tests, be sure to consider your market so that you can tailor each test to fit your audience. One way to do this is to build buyer personas that can help you find out who to target and what they’re looking for.

Then, choose which persona you want to target for a specific test.

Create two or more pieces of content that are similar enough that they convey the same underlying message, but different enough that you can measure them against one another. At the end of your test, check to see which version yielded the best results. That’s the version you should choose.

There are many A/B testing tools out there, but Freshmarketer is one of a kind.

A/B testing tools

Freshmarketer is a one-stop shop for optimizing and gaining insights about your website all from one place.

Plus, Freshmarketer’s Chrome Extension instantly loads your website with a live editor, so you can generate reports and track visitors right from your own pages.

Complete A/B tests with our editor and create all of the different variations of your site in just a few clicks.

It integrates with Google Universal Analytics so you can automatically import all of your reporting metrics into one location.

Split URL testing features can help you compare different pre-designed versions of your site, while heatmaps and scroll maps allow you to visualize user behavior and find out where they’re clicking.

With these tools, you can complete countless A/B tests to find which versions of your pages, site elements, and content bring in the most ROI.

One of the first things you should consider is your landing pages.

Landing page optimization

First impressions are everything.

In just 50 milliseconds (0.05 seconds), your visitors will form an opinion about your site that determines whether they’ll stay on the page or click off.

That’s why you want your site to engage users as soon as they click onto it.

Your layout, colors, content, fonts, buttons, and graphics should be optimized to make as many users convert as possible.

But you also shouldn’t have a one-size-fits-all page for every single visitor. You need separate, dedicated landing pages for visitors coming from different traffic sources. For example, you should greet users who head to your site from organic search with a different landing page than those who enter your site via email. That’s why you need to A/B test different variations of landing pages for different traffic sources.

Once you’ve created each variation, you can preview each one and track conversion rates for each of them all from your Freshmarketer dashboard. Once your test is complete, go with the landing page that yields the most conversions for each traffic source.

Your call to action (CTA) buttons can also affect your conversions.

Call to action buttons

If your CTA buttons aren’t click-worthy, then they aren’t living up to their full potential.

Just changing one simple button can have a significant impact on your conversions. Even adding or changing around the wording of your CTA can pay off. Just look at the results from the CTA button A/B test from Viadeo. They increased conversions on a page by 4.1% just by adding the words “It’s free!”

Try playing with the wording of your CTA, its location on the page, the style of the button, the size of it, it’s color, and more to find out which version has the highest conversion rate. Start with one variable and then test other changes over a period of time. When you finish testing, you’ll have a CTA that truly calls visitors to complete your desired action. Remember to A/B test segmented experiences to give different groups of visitors the most optimized version of your site.

A/B test segmented experiences

The personalization and targeted marketing options available today can help you gain a competitive edge against others in your industry.

By creating a customized experience for every visitor, you’ll see ROIs you’ve never seen before. Research shows that brands who use data-driven personalization see an increase in ROI of up to 20%.  

Segment and test your audience based on different factors such as:

  • Repeat customers vs. new customers: what messages will resonate best for new visitors and which ones are more suited for returning ones?
  • Keywords: are you using the correct keywords on your content? Think about search intent and possible search variations, like “designer shoes” versus “brand name shoes.”
  • Geography: are you testing different variations and messages based on a customer’s location?

Now that you know how and why A/B testing can improve the ROI of your marketing campaigns, you might be thinking, “How long should I run a test for?”

How long should I run an A/B test?

It might seem like a good idea to stop a test because it reaches some statistical significance.

But that’s too early. When you stop testing too soon, you’ll find that most of your lifts might have been imaginary. And that means little to no increase in ROI.

Instead, you should decide your sample size and run your test for at least two to four weeks before making any conclusions.

There are tons of tools out there to help you predetermine your sample size, like Evan Miller’s Sample Size Calculator. To use it, plug in the conversion rate of the existing version of the element you’re testing into the “Baseline conversion rate” box. This rate is the control.


Then, enter in the minimum uplift you want to see from your changes into the “Minimum detectable effect” box and press the “Relative” button. In this example, the minimum sample size needed per variation is 25,255 people for a minimum detectable effect of 5% with a baseline conversion rate of 20%.

That means you’d need to test the variation on at least 25,255 people before you could potentially reach your goal.


When it comes to marketing campaigns, who doesn’t want a bigger ROI?

A/B testing can help you achieve it by comparing different variables of your content, pages, CTAs, and more to find out which variable performs best. Setting up, running, and analyzing your tests is easier than ever before with Freshmarketer. You can integrate our tool with Google Analytics, make changes right from your dashboard, and view every test all from one central location.

Test different dedicated landing pages to find out which ones bring in the most conversions for audiences coming from specific traffic sources. Greet those coming to your site from email campaigns with a different landing page than those coming to your site from a PPC ad or organic search results.

Compare different CTA buttons to find out which colors, fonts, sizes, location, and wording are working best.  And be sure to A/B test segmented experiences so that returning visitors, new visitors, visitors with different search intents, and visitors in different locations see a personalized version of your content. Run your A/B test for a minimum of two to four weeks and predetermine your sample size with a sample size calculator tool to ensure that your analysis is as accurate as possible.

Start testing today to improve the ROI of your marketing campaigns. Even the smallest of changes can bring in some of the biggest results!