Ghost Buttons: The Good, The Bad, The Spooky: Part 2

ArticleNovember 17, 2015

Last month, we took an in-depth look at the ghost button, discussing its rise in popularity over the past couple years, the pros and cons, and when you should use one. 

We looked all over the web for examples of research or A/B tests to enlighten us if using ghost or standard buttons saw an increase in conversions. We couldn’t find any relevant data to answer this question, so our curiosity got the best of us, and we created our own A/B test.

Setting up the A/B Test

We set-up an A/B test using our monthly newsletter, which includes links to our most recent blogs and articles. Our email service provider (ESP) is Sharpspring, and we’re able to gather a great deal of data when we set up an A/B test using their email testing platform. 

Test A used our baseline newsletter template, which includes ghost buttons. Test B replaced these ghost CTAs with solid blue buttons. Everything else about the two versions was identical.

Ghost test emails

Test Results

After a week, we took a look at the Sharpspring metrics from the A/B test, and we found some eye-opening results.


The overall winner was Test B, the version with solid blue buttons. When comparing click rates, the solid button saw a 3.21% rate for clicks per deliveries and a 17.95% rate for clicks per opens. 

Test A, in comparison, had a 2.07% rate for clicks per deliveries, and an 11.11% rate for clicks per opens.

For every email opened, the solid button outperformed the ghost button by nearly 7%!

We were impressed that our opened emails with the solid button were clicked 18% of the time, so we went back and took a look at our average clicks per open rate for the last six months. 

Click rate graph

We were using the ghost buttons in all our newsletters before November, with click to open rate percentages ranging from 7.89% to 13.47%. In comparison, our solid button email, at 17.95%, performed an average of 11% better than our past 5 months of emails.

While this isn’t a true A/B test since the content and imagery differ month to month, these results do shed some light on the ghost button vs. solid button debate. It seems that, in the case of our newsletter anyway, solid buttons win time and time again.

What’s Next?

Given these results, we intend to replace all ghost buttons with solid buttons in our future emails. We will also explore replacing them on our website in the coming year.

Even though this single test showed ghost buttons were outperformed by solid buttons, we don’t assume this will be the case on every site or email. Ghost buttons can be strong, effective CTAs, so long as they are used wisely. We encourage everyone using a ghost button to test and find out for themselves if it’s creating the most conversions possible.