Should you split test your Facebook Ads?

Most marketing strategies will include split testing where two elements of your Facebook campaign will be tested against each other to see which will deliver you better results.

Applying this method goes beyond Facebook ads – it can be used for emails and landing pages that can result in possible improvements as high as 10x in ROI.

There are so many variables you can test but we have picked out some more common things you should try.

  • Call To Action Buttons
  • Images and/or videos
  • Headlines
  • Copy – long form and short form
  • Audience

It can be laborious, hence why so many people choose to skip it. However do that and you risk paying so much more for your ads and possible zero ROI.

Testing can take time as you want to do multiple testing, it can however get out of hand and add confusion so with the help of many guru’s, here’s what we know is worth testing.


  • Country
  • Gender
  • Interests
  • Age ranges
  • Custom Audiences

Ad Design

  • Image
  • Headlines
  • Text
  • Placements (where your ads are shown)


  • Ad types
  • Optimisation (conversions, clicks and engagement)

Once you have chosen some of these categories, set out to split test by starting broad and refine as you go.

What can possibly go wrong?

It’s advertising and let’s be honest, if you are about to go down this path and start split testing, be ready to lose some money.

Like any experiment, you will have some great wins and some not so great.  You need to run the ads for at least 3-4 days in order for the data to be reliable.

Don’t be so quick to shut a campaign down just because the cost of one variant was too expensive.  It can come down in time.

Facebook can produce some quick wins BUT you really need to be in this for the long term.  As you experiment you will learn what will work for your audience and your business.

Be patient!

Share the knowledge