Today, we’re talking about one of the most requested topics of our readers: A/B testing in Facebook Ads Manager.
So if you want to learn one of the smartest ways to make sure you don’t pay more for ads than you need to, join us on the inside.
In today’s post, we’re going to get into Facebook ad testing (A/B testing), including showing you how to do it!
A/B testing is an essential tool for any marketer or small business. It may seem like a bunch of extra steps at the beginning, but they’re worth it.
Let’s get into it.
- A/B testing helps marketers find the better ads so we can spend more money on those, improving our conversions and targeting.
- Running tests like this can also help you quantify and test your marketing strategy.
- Your team should be prepared to take the results and implement what you learned into your new marketing strategy.
What Is A/B Testing?
A/B testing is just a way to see which of the two options performs better.
It has been around for a while now, but we see it used often on websites or apps.
For marketers in 2022, we use A/B testing to test two components of a campaign or ad to see which will perform better with audiences.
A/B testing helps us find the better ads so we can spend more money on those, improving our conversions and targeting.
This saves you money from your social media budget in the long run and helps keep things like Ad Relevance metrics high.
Facebook takes early results and statistically calculates a winner with a confidence percentage.
Or, the chances are that similar results will be achieved if the test is repeated.
Facebook ad testing is the way to fast-track your ad account success.
And that is because it expedites your learning process, with Facebook doing most of the work for you.
Should You Use A/B testing For Your Ad Campaigns?
We shared some other reasons just a few seconds ago, but here’s a summary of why A/B testing is a good idea:
- Improves user engagement
- Increases conversion rates
- Decreases bounce rates
- Reduces cart abandonment
Running tests like this can also help you quantify and test your marketing strategy.
Maybe your Marketing Director thinks custom targeting will perform the same as interest-based targeting…
…but you’re sure custom audiences will be better for conversions.
If you don’t run A/B tests, you may never settle this dispute because you won’t have any data to back it up.
But if you test it, you can have results in just a couple of weeks that can guide you for the rest of your campaigns.
A/B testing isn’t just for graphics, font, or colors.
You can A/B test pretty much anything you need to. Let’s look at some examples:
Things You Can A/B Test In Your Campaign
- CTA (call-to-action)
- Product Sets
…and any combination of these and other things. Test whatever you want.
Before we talk about types of A/B testing and how to do the Facebook ad testing…
…we want to tell you to check out our new social media ads training course online.
If you’ve been marketing your own business or you want to learn how to become a digital marketer, these classes are for you.
The Most Common Type Of A/B Testing: Sequential Testing
You can get more advanced with your A/B testing by using sequential testing.
This is where you’ll run a series of tests to discover user preferences. Let’s look at an example.
You could run a test to decide what will perform better: a video or GIF?
Your initial A/B test tells you the video performs better. But now, you’re not sure if any video would perform better or if it was that specific video.
So you start A/B testing videos for different factors or elements.
Maybe your next test is two videos: one shot inside, one shot outside.
If outside wins, you can run another where you test a video shot outside with one person in it versus a video shot outside with a group of people in it.
You can get super specific with what you test here – font, colors, products shown, captions, and anything else that matters to your business.
How to Run a Basic A/B Test
You want to start simple when you start running A/B tests because there’s a bit of a learning curve.
However, A/B test setup and evaluation are not complicated at all. We will walk you through everything.
Here are the steps in the order we complete them:
- Decide what you want to test.
- Decide how you will evaluate the test’s performance.
- Create your control and your challenger ad in Facebook Ad Manager.
- Run the ad and monitor performance (optimize for the best-performing ad).
- Evaluate performance after ad ends… and repeat.
What Do You Want To Test?
For beginners, we usually recommend testing either audience (like targeting different audiences or showing different graphics).
Whatever you decide on, this will be the only difference between the two campaigns.
If you’re using Facebook’s built-in A/B testing, you’ll be able to easily make that change.
If you’re running an ad somewhere without A/B testing, make two versions of your ad and change the variable in your duplicate.
Use a Data-driven Approach to Goal Setting
Your A/B testing at the beginning may be a more stab-in-the-dark approach, but it should become a more scientific and honed process as you go.
The best tests are ones that start with a data-driven goal.
For example, let’s say most of your current buyers are women ages 21 to 28, but you think women in their 30’s would also like the products.
Your question is data-driven because you have the data that supports your buyer demos…
…and that data leads you to a question; will another age group perform similarly?
Run an A/B test with the same graphics (maybe change up the copy if you went hyper-young and cool, helps old folks out)…
…and don’t change anything else, except targeting.
Your A can be your regular targeting by age, group, and gender.
Your B group will change the age targeting to the next age range.
If you want to learn more about these ad targeting options, here’s our post on Facebook ads targeting that you can read next.
Now you need to figure out how you’re going to measure the success of your ad. Otherwise, what’s the point?
Establish KPIs (What will make your test a success or failure?)
If you don’t know how you’ll measure the success of your test, what are you running a test for?
You need to have a clear idea of your test goal and what a favorable outcome is.
Fortunately for us, most A/B tests already have a goal; find the best performing ad between these two.
And success happens when we get a winner or loser for the test.
If you get a winner, you have a clear answer to your data-driven question.
Your team should be prepared to take the results and implement what you learned into your new marketing strategy…
…whether that be a new-found audience or a surprise popular product line that needs to be enhanced.
A/B Test Best Practices
In a minute, we’re going to show you how to set up your A/B test in the Facebook Business Manager.
It’s a super quick process. But first, let’s talk about some Facebook ad testing best practices.
- There is only one major difference between your ad sets.
- Your ad sets have the same budgets.
- Your ad sets run for the full duration of the test.
Facebook Ad Testing: Setting Up Your A/B Test
You have a couple of options when you set up an A/B test on Facebook.
- You can set your ad versions up under the same ad set and compare results that way.
- Or, you can use Facebook’s built-in A/B testing.
If you choose the first option, you’ll just set your campaign and ad set up accordingly, and then create two versions of your ad under the same ad set.
If you want to use Facebook’s built-in option, you’ll do that while you’re creating your first ad version.
Let’s do that together now.
Facebook Ad Testing Using Facebook’s Built-In A/B Testing
If you want to set up your A/B test while you’re creating your ad, you can do that. Let us show you that process:
- Log into your Ad Account or Business Manager.
- Navigate to the campaign and ad set where your A/B test will be run.
If you’re creating all of these at the same time, click the green Create button in the campaign tab.
- Fill in the campaign and ad set details, if necessary.
- Name your ad and fill out details.
- Click Set up A/B Test and Facebook will automatically set up your B version. You’ll get to finish the ad when you hit Publish.
Using Experiments Area of Business Manager
We’ll show you how to do that.
- If you’re new to A/B testing, and you click into your Results tab, it will probably look something like this.
When you click Create New Test, you’ll jump to the Tests window.
- Click the blue Get Started button to start creating your test.
- You have two options for your A/B test; to Create a New Version or Compare Existing Ads.
We’re going to choose Compare Existing Ads but the processes are similar.
- Fill out test details. You’ll need to decide what you want to test; Campaign Groups, Campaigns, or Ad Sets.
The options A and B fields will auto-populate with a dropdown full of suggestions for you.
Select from these for your A/B test and pay attention to any errors or warning messages you get.
- Pick your start and end dates, and toggle on or off to end the test early if a winner is found.
- Enter your test name. We suggest making this descriptive so you know what it is at a glance.
It’s easy to find your stuff now, but when you have 200 A/B tests it’s not the same story.
- Enter the key metric you want to be measured and any extras.
You can add metrics by clicking the dropdown menu and making a selection. Remove metrics by clicking this X to delete.
- Check the estimated test power. Facebook recommends a test power of 80%.
Sometimes, like in our example, Facebook can’t determine an estimated test power. In that case, you’ll just have to go with it.
- Click Review Test Details (or fix any errors you need to fix).
You’ll also be able to see results from your A/B test in the Experiments Center. Let’s look at that now.
Viewing Facebook Ad Testing A/B Results in Experiments
Facebook segments A/B test results in the Experiments section in the Facebook Ad Manager.
This is what it will look like if you’ve never run A/B tests before.
- Log in to Business Manager.
- Use the hamburger menu to navigate to Experiments.
- The Experiments page will look like this if you haven’t used it before. You’ll see there are two tab options for you; results and test details.
- Results summarize performance for your chosen cost per result metric (like cost per purchase).
- Tests summarize your overall test and ad campaign activity and results.
This is where you’ll monitor the tests you’re running and view results.
You can also set up new A/B tests here and see Facebook’s suggestions for tests to run.
To view your A/B test results:
- Go to Experiments.
- Go to the Results tab.
- Find the A/B test you want to see results for and click View Report.
Understanding Facebook Ad Testing Results
When your first A/B test is done, how do you know what to make of the results? What would you look for as an account manager?
We’re going to walk you through test results, and what we look for when running A/B tests for clients.
Once a winner is determined, you’ll receive a notification in Ads Manager and an email will be sent to you with the results.
Let’s start with a “best-case scenario”.
a. The “Winning ad was found” statement is displayed.
If this is the case, Facebook was able to find a winning ad for you…
…so you’ll have a clear direction for your future campaigns and a good starting point for your next test.
If you’ve got a winner but you think your ad spend is higher than it should be, Facebook says that could be because of a few reasons:
- The length of the study was too short.
- There weren’t enough results to calculate an accurate winner.
- Best practices for study length weren’t followed or met.
You can work to lower the cost of your winning ad by following the best practices above.
Or, you can also check out our post on how to lower your Facebook ad costs next.
Now, let’s say you:
b. Don’t have a winner. Now what?
If you don’t have a winner from your test you may feel a little… deflated.
It’s annoying not to get a result from a test.
But it’s actually a common occurrence – not every test gets a clear winner. And that’s okay.
In fact, a study performed a couple of years ago showed that only 1 in 7 A/B tests gives you a result.
Usually, when you don’t get a winner, you choose a different variable (or make sure there’s more difference between variables) to test next time around.
That’s part of the fun of A/B testing.
Now if you want to maximize the results you’re getting from your Facebook ads, without spending so much time learning and doing A/B tests by yourself…
…consider working with our Facebook advertising agency.
We offer results-driven solutions for businesses like yours. Contact us today to get started!