Effective A/B Testing for Meta Ads

Maximizing Ad Performance: A Comprehensive Guide to A/B Testing on Meta Ads

In today’s digital age, advertising has evolved into a highly sophisticated art form. Gone are the days of placing static ads in newspapers and hoping for the best. With the advent of social media platforms like Meta (formerly Facebook), advertising has become a dynamic and data-driven endeavor. A/B testing is a crucial component of digital ads, in particular Meta ads, as they still give us good data on the variance in performance of different creative (unlike Google which would rather do the A/B testing for us in the background without really telling us what’s going on). In this guide, we’ll delve into the intricacies of building ad creatives for A/B testing on Meta Ads, with a focus on image and copy variations.

The Power of A/B Testing

Before we dive into the specifics of creating ad creatives, let’s emphasize why A/B testing is so important. A/B testing, also known as split testing, involves creating multiple variations of an ad and then comparing their performance to determine which one is the most effective. This data-driven approach allows us as advertisers to make informed decisions, optimize ad campaigns and should guide our content strategy moving forward, whether it’s colours, types of images, styles of wording or specific features that are highlighted in the copy.

Meta Ads, like many other platforms, provide advertisers with the tools to perform A/B testing easily. When done right, A/B testing can lead to significant improvements in click-through rates (CTR), cost per clicks (CPCs), and as a result, conversion rates, and overall return on investment (ROI) or cost per acquisition/lead (CPA)

When it comes to Meta ads, we have two main pieces of content we can test: the images we use in ads, and the copy (text) that accompanies those. This article is therefore written to explain how we can A/B test these two key types of content.

Building Images for Ads: Square and Stories Formats

One of the key aspects of creating effective ad creatives for Meta Ads is making sure that every image you test is able to be served in both news feeds and stories feeds. To achieve this every image variation that you are testing must be built in a square (1080×1080) and stories (900×1600) format.

One key consideration to keep in mind is that news feed (square) ads come accompanied with a primary text in IG, plus a headline on FB, so the image itself doesn’t need much copy on it. For stories ads, no other copy accompanies the ad (or it can, but it looks pretty terrible), so overlaying copy on top of the image, or separating it into an image and a copy block is important.

Newsfeed/Square Format Images

Square ads are known for their versatility and ability to grab attention. They are displayed seamlessly within users’ newsfeeds, ensuring that they don’t disrupt the browsing experience. When designing images for square ads, consider the following tips:

Visual Appeal: Use high-quality, eye-catching images that are relevant to your ad’s message. Clear, crisp visuals are essential to capture the viewer’s attention.

Branding: Incorporate your brand’s colors, logo, or other branding elements to maintain consistency across your ad creatives.

Call to Action (CTA): Can include a compelling CTA in your image to prompt users to take action, but it’s not totally necessary, as CTA buttons are added by Meta

You could just add a brand logo
You could add a tag line, a CTA button and a logo
Or just use the image on its own
Because ads come with headlines and primary text copy, you don’t actually need copy on the image for newsfeed ads

Stories Format Images

Stories ads are immersive and full-screen, making them a fantastic choice for engaging users. Whether it’s Instagram Stories or Facebook Stories, the principles for creating effective ad images remain similar:

Vertical Orientation: Stories ads are displayed vertically, so design your images accordingly. Use the full vertical space to captivate your audience.

Text Overlay: Your ad should contains a text overlay. Keep it concise and legible. Meta Ads have text overlay guidelines that you should adhere to for optimal performance.

Quick and Captivating: Stories are ephemeral and short-lived, so your ad needs to grab attention within seconds. Use captivating visuals and concise messaging. It’s worth noting that your messaging needs to be much more concise with a story ad, as there isn’t space for much copy.

Stories ads do not come with any other lines of copy, so messaging must be included within the image

Consistency Across Ad Variations

While it’s crucial to adapt your ad creatives to different formats like newsfeed and stories ads, maintaining consistency in your messaging and branding is equally important for your A/B test to be effective. Consistency helps build trust with your audience and reinforces your brand identity. Here are some key points to consider:

Unified Branding: Ensure that your brand’s visual identity, such as colors, fonts, and logo, remains consistent across all ad variations. This helps users recognize your brand instantly.

Message Alignment: Your ad copy and imagery should align closely with your campaign’s objectives. If you’re promoting a specific product or offer, make sure this is clear in all ad variations.

A/B Testing Image Variations

Now that we’ve covered the importance of format and consistency, let’s delve into the specifics of A/B testing image variations. Meta Ads provide an excellent platform for testing multiple image variations to identify what works best. Here’s how to approach it:

Create 2-3 Image Variations: Start by designing 2-3 different images that convey your message effectively. These variations should differ slightly in terms of visuals, style, colours or focal points.

Testing Parameters: Set clear parameters for your A/B test. Decide on the metrics you’ll be monitoring, such as CTR, conversion rate, or CPA.

A/B Testing Copy Variations

A/B testing copy variations is equally essential as testing images. The text in your ad plays a significant role in conveying your message and persuading users to take action. Here’s how to approach A/B testing for ad copy:

Create 2-3 Copy Variations: Craft 2-3 different versions of your ad copy. These variations could  convey the same message but with slight differences in wording, tone, or emphasis. Or, if you have different USPs or reasons people might buy/sign up for your service/product, then test different one to highlight. Do you highlight price or features? Speak in a data-focused way, or a more emotional way? These are all things you can test and iterate on.

A/B Test Alongside Images: Run your copy variations concurrently with your image variations to assess how different combinations perform together. Make sure that every image variation is tested with every copy variation. If this doesn’t happen, then you won’t know that it was the copy making the ad perform better than the image, or vice versa.

Ad Positioning: Consider the placement of your copy within the ad. Whether it’s in the headline, body text, or CTA, test different positioning strategies.

Analyze and Optimize

Once the test is complete, analyze the data to determine which image variation performed the best. Pause the poorest-performers, and build new ads in a style more similar to the winners. The aim here is to keep rotating ads (which you should do anyway to prevent ad fatigue), always pausing the worst and creating new that are better. In this way, your performance overall should improve with time.

Testing Duration: Run the A/B test for a sufficient duration to gather statistically significant data. Avoid making hasty conclusions based on short-term results. Really, duration is irrelevant, what you need to have good statistical significance is either impressions (if you’re testing for CTR) or clicks (if you’re testing for conversion rate/CPA. At least 5,000 impressions per ad is important to establish better or worse CTR. At least 200 clicks per ad should be enough to establish better or worse conversion rates. However, if results are very similar, you may need to continue the test.

Remember that if results are similar and one or two conversions would shift the results from one variation to the other, you don’t have enough data to be a reliable predictor of future performance.

The Potential for Many Ad Variations

By combining three image variations with three copy variations, you can create nine different ad variations. This allows you to explore a wide range of possibilities and gain valuable insights into what resonates most with your audience.

Here’s how these variations can be structured:

Image 1 + Copy 1

Image 1 + Copy 2

Image 1 + Copy 3

Image 2 + Copy 1

Image 2 + Copy 2

Image 2 + Copy 3

Image 3 + Copy 1

Image 3 + Copy 2

Image 3 + Copy 3

This comprehensive testing approach allows you to fine-tune your ad creatives for optimal performance, ensuring that you’re delivering the right message to the right audience.

Remember that each ad needs 5,000 impressions (for CTR testing) or 200 clicks (for conversion testing). As such, if your budget is on the smaller side, you may need to keep your test on the smaller side to be able to get results sooner rather than later.


A/B testing is a powerful tool that enables us as advertisers to refine our ad creatives for maximum impact. I would say that when given the option, I usually opt to test images first, as I think the difference in how eye-catching they are can tend to have more effect than the ad copy, bit it all depends on what is being advertised. Test having copy overlaid on the image, or just using the image alone. This can have quite a large effect. Perhaps you want to just use the image with a brand logo in it to look a little more polished. Most importantly, use eye-catching images! These can be your own photos, they can be made by graphic designers, or even try AI images. All can yield quite different results. But above all, enjoy the analysis, and try to put yourself inside the mind of your audience to figure out why they like one more than the other.

As a side-note, I have often seen that an ad with better CTR ends up with a lower conversion rate. I have theorised that this is because the ad is so enticing that it attracts people that aren’t actually that engaged, and don’t end up converting. A lower CTR might mean it doesn’t appeal to such a broad audience, but the people that do click on it know what they want, and are therefore more likely to convert. Is CTR more important than conversion rate, or vice versa? Looking at CPA is helpful, and if they end up producing a similar CPA, I will always choose the ad that has better CTR, but as with everything, it depends. Your own analysis is what is really important here.

Leave a Reply

Your email address will not be published.