In 2015, AdEspresso analyzed 37,259 Facebook ads and found that most companies only have one running ad while the best use tens or even hundreds of different ads.
Why would a company run hundreds of ads at once?
After all, the potential likes and shares will be scattered between the ads.
But the truth is, the value of A/B testing your ads outweighs the potential likes/shares underneath the post. There is no greater value for a marketer than actually grasping their audience’s interests, preferences, and personalities.
Are you ready to uncover the secret desires and psychological peculiarities of your audience?
To answer all the burning questions you might have about your target audience, we’ll explore ten Facebook ad A/B tests and discuss the best practices.
Here’s to all of the unlimited opportunities of split testing!
The essentials of A/B testing
Let’s start with a short-but-sweet introduction to split testing (scroll ahead if you’re already familiar with the concept).
Put simply, the Facebook ad A/B testing involves creating multiple variations of a similar ad, with various images, headlines, texts, call-to-action buttons, etc. You can also split test your audience specifications, such as the location, interests, gender, age, and much more.
The Golden Rule of split testing (that many marketers forget) is to find out which ad combinations help to achieve your END GOAL.
For example, if you sell online software and your end goal is to get new users to sign up, monitoring cost-per-click isn’t as relevant as discovering the ads with the highest conversion rate.
To grasp the amazingness of A/B testing, here’s a graph by AdEspresso that explains all the opportunities.
As you can see, even a few small alterations will result in 45 different ad combinations. And this is exactly what makes the split testing process so much fun –you’ll be able to compare the smallest differences and discover what works.
By the way. Creating tens or hundreds of ad variations doesn’t have to be daunting.
No, you do not have to create all of those ads manually. I like using AdEspresso to set up all our A/B tests. It is a quick and straightforward process and in 3 minutes, I managed to create 32 ad variations. Easy peasy!
Moreover, you can use AdEspresso to optimize and analyze your campaign performance. Just to make sure you’ll get some value and insights out of the tests. (Don’t fall into the lazy-person trap of creating a set of Facebook ads and never analyzing them! You’ll have to start all over the next time.)
In case you’d like to read more about split testing on Facebook, here’s the evergreen guide to Facebook ad optimization and A/B testing.
Now I bet you’re all wondering what ARE these questions your can answer with split testing.
Well, well, well… Alright, I’ll tell you.
1. The Crown Jewel – What’s your most compelling value offer?
Mirror, mirror, on the wall, who’s the fairest of them all?
Naturally, we would all like the mirror to point on our excellent products. But what if our product has many faces?
I envy the companies who have discovered their most effective unique value proposition (UVP) and continually use it in their marketing campaigns. For the rest of us, it is a neverending struggle to find the one and only value offer that works the best.
Do you know what’s your products’ greatest value offer? No worries, you’ll get closer to the right answer by A/B testing your Facebook ads.
To find the most compelling value offer, all you have to do is to create ad campaigns with all of your value offers (even better if you also create separate landing pages for each UVP.
See how CoPromote has done it:
In the first ad, their unique value proposition is “Reach 500 000 new people per month”. In the second ad, they use the UVP of “Cross-promote with creators on Twitter, Tumblr, …”
By using two different value offers, CoPromote can see which ads have the highest click-through and conversion rates.
This tells them two things:
- Which offer has a higher click-through rate – what offer makes people click on the ads
- Which UVP returns a higher conversion rate – which offer is better-targeted and makes people sign up
Naturally, the goal is to discover which offer makes people sign up for the service.
By split testing your value propositions in the ad’s headline or text, you’ll be able to discover real gems as well as test the new approaches to your product before using these on the grand scale.
2. Mobile or old-school Desktop Ads – Which convert on higher rates?
In the fourth quarter of 2015, Facebook reaped 80% of its ad revenue from smartphones and tablets. The mobile ad industry is HUGE and growing still.
But are Mobile ads always the best choice for your company?
The answer lies somewhere between “yes” and “no”. And the best way to find out is to conduct a simple A/B test.
A/B testing mobile vs. desktop placement is one of the several split tests that you can do with AdEspresso (in a snap). After you’ve entered your campaign details, the system will ask you whether you’d like to test additional parameters. Desktop vs. Mobile is a fairly simple test as it only doubles the number of your ads (while testing interests can create a much bigger ad set).
As you can see from the image below, ad placement can make a huge difference in your click-through-rates and the cost-per-click.
But remember (and I’m going to say it again):
Collecting clicks is not your end goal. So monitor which ad placement has the lowest cost-per-conversion, the least expensive cost-per-download, etc.
The efficiency of mobile ads depends on your offer. If you want people to read your blog articles, a mobile ad is The Guy.
But if you’re like Marketo, and want people to download an eBook, mobile ads are not the best choice – try reading an ebook on mobile, it’s a time-consuming and irritating task. And it is highly likely that people seeing your ad do not have a wi-fi connection to download the eBook.
Before you begin A/B testing your ad’s placement, double check your mobile advertising best practices checklist: make sure you have a mobile-friendly landing page, that your ad text is short but sweet, and that people can clearly understand what you’re offering to them.
3. What happens after the click – Which landing pages perform the best?
Unless you’re using Facebook ads that do not require a landing page (e.g. lead ads), you need to optimize your landing pages for high conversions.
As we started split testing our ads in my company, Scoro, I had many burning questions demanding for an answer.
Here’s what we wanted to know:
- Should we lead people to our home page or our features page?
- Should the landing page open up with a sign-up box or without it?
- Are industry-specific landing pages more effective than using our home page?
After running some tests, we learned that the landing page showcasing our product feature converted 24% more visitors than our home page. We ended up saving hundreds of dollars.
You can even test your landing page layout and visuals by A/B testing Facebook ads.
For example, you could check whether a colorful yet simple landing page is a better option than using a background with people.
We’ll never know which of the following landing pages have a higher conversion rate, but make your guess:
The simple page design by PhotoKeeper…
Or the landing page by Hubspot that uses people as background?
We don’t know the performance specs of these two sites, BUT you can totally test what works for YOUR company.
Testing your landing page performance is a super important part of Facebook advertising. If you’d like to learn more about creating good landing pages, start by reading this guide on building high-conversion landing pages.
4. Ice cream, marketing or architecture – What is your audience interested in?
Name a marketer that doesn’t care about their audience’s thoughts, interests, and aspirations? I could come up with none.
We’d all like to know our audience better. And this is yet another moment when split testing enters the game.
By using AdEspresso, you can quickly set up an ad campaign and track what your audience is interested in.
The main benefit of split testing your target group’s interests is that the next time you’ll know exactly whom to target.
5. London vs. New York – Where is your most engaged audience located?
If you run an international business, you need to know where’s your most engaged target market. But this isn’t the only location-related question that you can answer with A/B testing.
In fact, there’s a whole bunch of different cues about your audiences that you can easily collect.
Not all ads work the same way in all countries. As we advertised in multiple countries, our red-tone ads worked the best in the Eastern Europe countries. In the UK, simple and light ads stole the show.
Here’s how to get more out of split testing your ads:
- First, create an international ad campaign with various ad images and texts.
- Next, see which ads performed the best in each target country and use this data to create new ad campaigns for each location.
This time, you’ll know what works and can bet all your money on the winning strike.
6. Picture-perfect – Which colors work the best for your ads?
I’m a big believer in color psychology. All things considered, I think that nobody can overestimate the importance of colors when it comes to Facebook advertising.
Various colors symbolize different emotions, affecting the viewers in unique ways.
To know which colors work the best for your audience, conduct an A/B test with various ad visuals.
Take a look at these four ads by The New York Times:
By running ads with different color schemes, the company can attract more people’s eyes while battling the ad fatigue. Moreover, they’ll learn which tones are the most engaging when shown to their target audience.
Use the color psychology to your advantage and learn the emotions and feelings that each tone signifies. You’ll be surprised by how big a difference another background tone can make.
7. A human touch – Should you use images of people or pictures of your product?
Many marketers suggest that we use people on ad visuals – it will create a stronger connection between the brand and the audience and generate more trust.
But one can never be sure, and so it’s necessary to test both images WITH people and WITHOUT.
We can only guess which one of these ads by Plated had the highest conversion rate:
- A single image with people
- Two images: one with people and one with products
- A single image with products
Ads visuals are one of the easiest things to A/B test. They’re also the most important element of your ad, delivering the first impression of your offer.
8. To text or not to text? – Should you fill in all text boxes?
When creating ads for your Facebook campaign, do you fill in all the fields?
After testing hundreds of different Facebook ad combinations, we’ve discovered that sometimes, the best message is using no text at all.
Creating ads with short or no text requires a strong ad visual, telling the whole story. If you’ve managed to come up with a particularly great idea for an ad image, testing mobile ads with and without text might be an interesting experiment.
If you’re unsure about using no text at all, start by split testing the length of your ad messages.
Which one do you think worked better:
The short yet straightforward text in Quip’s ad…
Or the longer explanatory introduction in this ad by Greenhouse Software?
9. Download, Learn or Buy? – What’s the best call-to-action for your ads?
Not a long ago, I wrote a blog article about testing over 100 Facebook ads in one month.
One of the most notable findings was that “Learn More” and “Sign up” call to actions outperformed each other in different situations.
In our case, the “Learn More” call-to-action button managed to attract a higher click-through rate.
But… Wait for it… The “Sign Up” CTA returned a higher conversion rate!!
So which one in the Facebook gallery of call-to-actions should you pick?
The next time you create a Facebook ad campaign, test multiple call-to-actions to discover the most effective one and optimize your ads accordingly.
10. Are you being too kind – Should or shouldn’t you offer something for free?
During your career in advertising or marketing, you’ve probably created at least one free giveaway. That’s because offering something of value for free instantly catches people’s attention and helps to increase brand awareness.
But what if you’re interested in getting new customers? Are free giveaways the right way to go?
Once again, this is a question that you can answer by A/B testing your ads and analyzing the results one month after the campaign’s end.
Here’s how to test the efficiency of free giveaways vs. a paid offer:
Step 1: Create 2 campaigns: one with an entirely free offer and another that requires a small commitment. For example, take a look at this New York Times campaign offering 4-week free readership for $0.99.
Step 2: Measure each campaign’s performance, collecting data about the click-through rates and the number of sign-ups.
Step 3: (And this is the most important part!): One month after your campaign has ended, track how many people started to pay for your service.
Moreover, calculate which one of the campaigns delivered the lowest cost-per-acquisition. Only then can you truly evaluate the performance of your two ad campaigns.
This is an extremely valuable split testing method, as you’ll see whether a small commitment, in the beginning, made people more likely to stay with you in the long term.
I hope you’ve gathered some interesting insights and ideas for A/B testing your Facebook ads. Whichever questions you have, A/B testing is likely to provide the answer.
As our CEO Massimo always says:
Never assume anything. Always test everything.”
Don’t forget to share your experience and split testing ideas in the comments section. We’re eager to find out! 😉