Facebook is unquestionably the world’s largest social media channel. Studies indicate there are approximately 1.23 billion people interacting with Facebook content on a daily basis, and 1.15 billion people use mobile devices to get their content fix.
Obviously, brands large and small eagerly attempt to build connections with this massive pool of prospective customers.
However, when it comes to actively advertising your brand on Facebook, there are a number of best practices that you should follow. Running ads for the sake of simply promoting your brand is an ineffective tactical approach to Facebook.
You need a plan to make those ads work for you and build connections with the people who see your content. To build a sustainable plan for Facebook advertising, you need to A/B test the types of ads you promote across the channel.
How should you plan to A/B test Facebook ads? Let’s look at a few best practices to help you build an effective model.
1. Determine a Workable Hypothesis
A strong hypothesis is the foundation of your A/B test. Your primary concern with Facebook ad testing is to learn what types of content resonate with your audience and what types are less successful.
Once you have enough data from your test results, you can reasonably prove or disprove your hypothesis.
How to write a strong hypothesis for Facebook
Your hypothesis correlates with a strategy, which must outline how to solve an existing problem. The basic structure of a good hypothesis involves 3 main points:
What is the problem you’re trying to solve?
What is the solution or change you presume will solve the problem?
What is the anticipated outcome your test will have on your key performance indicator?
Document your hypothesis in a way that alludes to the problem without wedging it into the statement. For example, if you identify that stock or recycled content is turning people away from your ads, write down your hypothesis so that you highlight the anticipated impact of your solution.
Remember that your goal with A/B testing is to learn
The test itself leads to your success. Even if the data you collect from the results disproves your hypothesis, you still learn from the experience.
You can take the results back to your team and brainstorm a new hypothesis to test against Facebook ads. Continue this process until your test results indicate that a change to your content or your positioning will improve your KPIs.
One thing that all marketers can agree on without the need for A/B testing is this - continuous learning is always a good thing.
2. Identify a Sufficient Audience to Test
To determine whether your content or your messaging is effective, you need a data sample that’s large enough to draw reasonable conclusions. Targeting your ads towards an audience that’s too small or unaligned with your ideal clientele will skew the data, and ultimately make your test ineffective.
If you’re conducting large campaigns then you’ll want to use Facebook Ad Manager or Power Editor, which are designed to help large brands measure multiple campaigns at once. Facebook created how-to guides on using both Ad Manager and Power Editor to help you ramp up on the two tools.
You can also use these tools, especially Power Editor, to target people who follow both your Facebook and Instagram channels. Ever since Facebook acquired Instagram 5 years ago, the two platforms have done an admirable job of integrating their features and audiences to help marketers connect on every possible level.
Determine basic demographics
Select an audience that is appropriate for your test. You can choose to target your existing followers, people who follow your followers, or a brand new pool of prospects with little to no interaction with your social content.
You’ll want to decide on:
The appropriate age of your ideal user
How (or if) to test by gender
What geographic location the ad will appear
Any related content that intrigues your target demographic
Whether annual income matters and how much is your benchmark
Target by psychographics to find the right types of people
This is an important part of your audience profile, especially if your KPI is clicks. You want the right kinds of people to interact with your content, preferably ones with potential to become paying customers. Targeting based on characteristics like interests and behavior will help you reach people aligned with your ideal user base.
There are many different filters you can apply to your Facebook ads, which can be done on an as needed basis. Keep in mind that if you’re conducting an A/B test, you want to keep as many variables as possible consistent to draw reasonable conclusions from your test.
If you want very specific filters in place for one of your ads, you’ll need to apply those same filters to every ad within your A/B test for accurate results.
3. Test Against the Right KPI
You need to decide which KPI your test will be measured against, which you’ll identify within your hypothesis. This is the blueprint for determining the effectiveness of your Facebook ads.
Will you measure click-through rate (CTR)?
Clicks on your ads are one of the most common ways to measure Facebook A/B testing. According to data compiled by Amplified, the average Facebook ad CTR across all industries is 0.93%.
You can see in the chart above what is the average CTR across various industries. Use those numbers as your benchmark when testing the CTR of your Facebook ads to see how you measure up against the industry average.
If your data is on point or above the industry averages cited above, you can conclude that your test was a success. On the other hand, if your CTR comes in well below the industry average, it’s time to go back and rethink your content or possibly your entire campaign.
Will you measure conversion rate?
If your goal is to go beyond clicks and early interactions with your ads, you’ll want to analyze the lift in conversions on your website. One way to measure Facebook ads is to analyze how people take action on your website via form fills, views on a particular page, and requests for demos.
Conduct an A/B test to determine which types of ads inspire the right kinds of people to interact more with your site. Establish a filtered segment in Google Analytics or whatever analytics tool you use to measure the flow of traffic on your website to see what kind of impact your ads have on website conversions.
A positive lift equals a successful test.
Will you measure frequency, spend, and return on ad spend (ROAS)?
Neil Patel offers some helpful insights for analyzing Facebook metrics. Should you test your ad campaigns by frequency, you’ll want positive results in a very short period of time.
You can also measure the revenue you earn from your ads against the amount that you spend to run the campaigns. Analyzing revenue alone is ineffective as your cost per campaign, particularly if you’re running multiple campaigns with hundreds or thousands of dollars in ad spend, takes money away from the revenue you earn per converted prospect.
Think about what metric is the best option for your vertical and your brand before you begin A/B testing. Remember to select ONE metric to measure against your campaign(s). If you try to calculate the value based on too many variables, the test will spin off its wheels and ultimately prove ineffective.
4. Select a Timeline and Stick with It
Your A/B test needs time to run before you can draw any logical conclusions. This means that once you decide on your hypothesis, your KPIs, and your audience segment, the ad needs to run for a reasonable amount of time. Once it begins to run, you need to wait until time runs out.
Average time to run an A/B test
According to Facebook’s best practices guide, the optimal run time for A/B tests is between 3 and 14 days. That being said, the length of time that’s best for your brand and specifically your campaign is contingent on a few factors.
What’s the size of your test audience? If you’re going after a large pool of prospects, you need to give your ads time to reach enough people and collect enough data.
Is there any seasonality or special promotions that influence the timing of your content? Does that influence the amount of time you have to conduct A/B testing?
How much budget are you setting aside for Facebook ads? The size of your budget should influence the amount of time and capital you spend testing your Facebook content.
Stick to the same test run for new content
If your ad has come to an end and you’ve collected enough data to warrant changes to the content, go right ahead and do what the data tells you to do. But when you test your updated content in a new ad, you’ll want to run it for the same period as your previous ad.
Keep the timing of the ad run consistent with your previous content, and keep the other variables - audience, ad spend, etc. - equal to your past test. This way you can directly analyze the impact of your change, and make better decisions with your content moving forward.
Facebook is a critical part of any marketer’s plans for content. As the largest social media channel, any plans for content on digital channels need to run through Facebook. Register for our upcoming “How to Create Facebook Content with Impact” webinar to learn more about evolving consumer behavior, creative considerations for Facebook content, best practices for A/B testing, and how to use the tools that help you achieve your goals.