Testing. Testing. 1-2-3. Is this thing on? Do you want to test your ads but don’t know where to begin? It can be a daunting task, and you will have a lot of questions. Once you lay out your plan and follow a few rules, you can get a lot of data to apply to your digital marketing strategies. In this article, we will focus on Facebook advertising. You can also check out this blog about Google Ads experiments.
Facebook A/B Testing
Use this method to test one variable (such as two versions of your videos) served to an audience that is equally split with the same budget. You might also know this as “split-run” testing. Facebook Ads Manager has a great feature that walks you through the steps to create an A/B Test. You can also create the Facebook A/B Test in Experiments by selecting an ad to test it against its duplicate version or picking two existing ads and comparing them.
Now that you know where to perform the Facebook A/B test, here are some ideas of what you can test.
Creative (Includes Copy)
We recommend always testing your creative (even if it’s not as a formal A/B test) since it is one of the proven advertiser controls that can positively affect your account performance.
- Lifestyle photos (or environmental shots) vs. product images
- Static image with text overlay vs. static image without text overlay
- Different versions of videos (test lengths or videos with/without captioning)
- Long copy vs. short copy
Know your audience! You can learn which ads resonate with your different audiences. Will some of them convert higher with lifestyle imagery? Can you achieve a lower CPA with existing audiences?
- Core audience vs. lookalike audience
- Broad audience vs. custom audience
We know Facebook has sophisticated machine learning, and it’s best to leave it to do its thing. But we can also benefit from testing it. Try placement testing if you have specific creative (such as Stories), and choose manual placements.
- Manual vs. automatic placement
This is an overlooked variable to test that can give you insights into your customers’ behavior. Try testing where you are sending your ads. Test sending your ads to the homepage and to the category or product page.
- Homepage vs. category or product page
What Else Do You Need To Get Your Tests Ready?
Invest in your experiment with a good size budget. You don’t want to skimp on this and not reach statistical significance. Consider using a daily budget and letting your ads run at the same time.
Determining the Winner
Before you begin your test, think of what the measure of success will be. Is it cost per click, or will it be conversion rate, cost per conversion, or cost per 1,000 impressions (CPM)?
At the conclusion of your experiment, pull together your data and theories. Your agency can help you view and understand your A/B test results and guide you with a strategy and an action plan. Remember that one test does provide valuable insights, but major business decisions should not be made solely because of them. Continue to test if you want to prove your hypothesis, and adjust your expectations accordingly.
There are other tests that can be performed when you have agency support and contacts. Agencies like JumpFly have several clients participate in Facebook testing opportunities that may be partially and even fully funded. Talk to your agency if you are interested in participating.