The Ultimate Guide How To Test Facebook Ad Creatives in 2024

Everyone knows how important it is to test new ads

Still, the most common mistake I see scaling brands make is doing wrong.

They either test random stuff

Add their new ads to campaigns that are already running

or simply not doing enough of it.

The difference between testing correctly and incorrectly is thousands $$$ wasted.

And more thousands, potentially millions $$$, lost in an opportunity.

While the “meta” constantly changes, testing will remain foundational in every digital marketing strategy.

Here is an example before and after improving the testing:

Okay, testing is essential. How do we test stuff?

Let’s get started!

General rules when testing.

The stricter these are, the easier the decision-making process is.

However, sometimes these could be a bit more “soft” depending on other factors we will discuss at the end.

Step.1 - Decide What you want to test.

Usually, this is:

• Creative

• Copy

• Audience

• Headline

• Element of these (For example, The hook of the copy or the creative if it is video)

⚠️ How to decide what you want to test. The whole premise of testing is to make data-driven conclusions, including what is to be tested. By analyzing what is happening in the ad account, we can decide what could have the biggest positive impact. Generally, this is the creatives, but this is not always true.

Step.2 The Structure.

Setting up the campaign is easy. You go with ABO (ABO is ad-set budget optimization). You select how much you want to spend per day on ad set level. CBO (Campaign Budget Optimisation), now named “Advantage+ budget,” remains off.

Why do we do this?

We want to use the algorithm with enough freedom to test the ads but not too much freedom.

So we want to confine it in a space and have more control over how much and where the budget is spent.

Using ABO (Ad Budget Optimization), we can do precisely this.

If you go with “advantage+” and have one ad set, it doesn’t make any difference, but you will have to create a new campaign each time, and generally, this setting is better suited when scaling.

It looks something like this:

Step 3 - Budgets & Testing

First, how much do you spend on testing? The general rule is to spend around 15-20% of your total budget on testing. If you increase this number, your margins may suffer and have lower profitability. At the same time, if you go too low, you may find yourself in a situation where your ads are fatigued without finding winners to replace them with.

We covered how much you want to spend as a percentage of your total budget, how much you pay per ad, and how to decide how much to put in as a daily budget.

This depends on a few factors like your AOV, CPA, and so on; luckily, there is a relatively easy way to calculate this.

Facebook recommends 50 weekly sales, which is often impossible for smaller brands with limited budgets.

1. When to pause ads Ads

We choose budgets in the ad set settings. You usually want 3-4 ads per ad set. If there are no sales, we want to pause ads at 2 x tCPA (target cost per acquisition).

Example: Your products cost $30, and your target CPA (cost per acquisition) is $15, so if you spend $30 and still do not get any sales, pause the ad. Brands with more expensive products likely would need more money for testing their ads.

!Disclaimer: In some cases, you may not get sales, but still, it is worth to running the ads a bit longer if the vanity metrics like (ATC, CPC, CTR, etc..) are too good.

2. When to pause ad sets.

If you pause your ads when they reach 2 x tCPA at one point, you will run out of ads in the ad set or have some winner. In the first situation, pausing the whole ad set at 4 x CPA is safe if none of the ads has generated a purchase.

Some ads may not be clear winners; they can generate sales but not hit the target numbers. In this case, run them longer; if you don’t see improvement, you can pause the ad. However, this gives us an indication that this ad could potentially be a winner with some tweaks.

3. What do you do with winners?

We can safely declare a winner in an ad that has generated 10 sales, hitting your target CPA. From here, we have many options:

1. Scale the ad set in the testing campaign

2. Pull out the ad ID and launch the ads in a separate scaling campaign

3. Test the ad on new audience

4. Test the creative with new headlines

5. Test the creative with new copy

You can go creative, but generally, these 5 will have biggest impact and are my go to.

Tip: Don’t pause stuff that is working. I see these mistakes exceptionally often; relaunching a campaign without any changes can give a completely different result.

Step 4 - Metrics

Here are some key metrics I always look at to evaluate the performance of ad creatives:

1. Clickthrough Rate (CTR)
   This metric shows how well your ads are performing. Good CTR varies from industry to industry, as the product, and so on. I like to use the unique link click CTR, which is a version of the CTR. Counts only the unique link clicks and is more accurate

2. Cost per 1000 Impressions or Cost per Mile (CPM)
   This metric indicates how well you are performing compared to the overall competition. It also dictates how much you pay 1000 people to see your ad. If ads are bad, you will have higher CPMs, and pay more money for advertising.

3. Cost per Click (CPC)
   A higher CPC suggests that the ad isn't performing well. The CPC is based on the CPM and the CTR, so if your CPM and CTR are in check, the CPC will also be good.

4. Conversion Rate (CVR)
   CVR measures the effectiveness of your ads and the platform in driving desired actions. You want to have this as high as possible, as it is one of the most impactful metrics.

   1000 visits x 2% = 20 sales.

   1000 visits x 4% = 40 sales

5. Return on Ad Spend (ROAS)
   ROAS is calculated by dividing the revenue attributed to your ad campaign by the cost of that campaign. It's critical to understand the profitability of our spending.

6. Hook Rate
   This is calculated as 3-second video views divided by impressions. It's a basic metric, but it helps quickly identify which ads are grabbing users' attention. Achieving above 30% for prospecting campaigns indicates creative solid performance.

7. Hold Rate (HR)
   HR tells you if the video was compelling enough to retain viewers until the end. Achieving above 10% for prospecting campaigns means the creative is effective. Tip: Keep surprising or asking questions in the middle of the video to retain viewers.

8. Landing Page View Rate (custom metric)
   This metric is calculated by dividing unique clickthrough rates by landing page views. It measures how many people who clicked on the ad visited the landing page. A rate above 70% indicates that the ad generates enough curiosity to drive action.

9. Frequency and Impressions
   These metrics go together. Impression measures how many people saw the ad, while Frequency measures how often they saw it. The ideal frequency should be between 1 and 3 per ad set. High frequency indicates that the ads may need replacement or expanding the targeted audience.

10. Cost per Action (CPA)
    CPA, which includes purchases, add-to-cart actions, or leads, is the main lever to see if the creatives are good.
    

Step 5 - Common questions & Mistakes

1. Do you use Meta's A/B testing feature for creative testing? Normally, no. However, when you want to test a slight change and show the same ad, you can always try.

2. How do you use dynamic ads for creative testing? This is type of test I use sometimes. I usually make this test in CBO campaigns and duplicate the ad set 3 to 4 times. The dynamic ads remain the same. After I ran it for a few days, I saw the trend in the breakdowns in the ads manager. However, often you will ahve to go through the standard ABO testing again.

3. Pausing ads too early.

4. Making changes to ads that have already been running

5. Turning ads on late in the day. This causes Meta to spend the entire budget in a few hours, often causing bad traffic and no sales.

6. Testing ads in the scaling campaigns. You want to keep these separate. Running them together will ruin your entire performance, as Facebook often prioritizes new ads, and testing ads will get a big chunk of the spending.

   
   
   

If you like this type of content and want to learn more about advertising and digital marketing make sure to subscribe.