Much of my work on Facebook ads involves split-testing. I split-test campaign objectives, ad copy, designs, audience targeting and more in order to drive new leads.
Running such split-tests often leads to surprising results. Ideas that we assume wouldn't work, can end up performing quite well. And vice versa, which makes these tests so valuable.
One recent example of a surprising split-test involved various calls to action in the copy. The client, a healthcare professional, wanted to acquire new patients and we split-tested 4 Different Calls To Action (CTAs) in the ad copy for that purpose.
The Calls To Action Tested
The first two CTAs invited users to
"Visit our website" or
"Click here"
People who clicked on the ad were taken to a landing page encouraging them to contact the office to book an appointment.
The other two CTAs were more explicit. They read:
"Click Here To Book an Appointment" and
"Click Here & Contact Us Today"
People who clicked were taken to the same landing page as above.
Finally, we ran one version of the ad that did not contain any Call To Action.
Guess which CTAs performed best? Our assumption was that the more explicit copy would lead to more appointments than the first two, and that the ad with no CTA would perform the worst.
Results
We discovered that visitors who came from the ad without a CTA converted nearly 10% more than those with a CTA!
There was little variance in conversion rates between the other CTAs.
In other words, people who clicked on the ad without the CTA were 10% more likely to call the office or submit a contact form than those who were explicitly told to do so.
This may have been a one-off fluke, so I tested it on different audiences within certain radii of his office, and the ads without the CTA still outbeat the ones with a CTA, albeit by a smaller percentage.
This goes against the grain of marketing best practices, where it's been drilled into our heads that we need to tell the user exactly what to do at the first opportunity.
It still may indeed be a best practice and no one should stop using CTAs based on one client's results during a limited time frame. But I would argue it's an even greater best practice to challenge these notions. Keep on testing!
Commentaires