Advertising online can sometimes be intimidating. There’s simply so many ways to approach it. How could you ever know if you’re doing it right? And even though an ad campaign being limited to Facebook or Instagram helps cut out some of the overwhelming outside factors, there’s still so many options for your ads!
That’s where A/B testing comes in.
Facebook Ads (And Their Instagram Counterparts)
Since it’s the 21st century and you’re a human being who’s probably older than 20 or so, chances are you’re on Facebook. And since you’re likely on Facebook (or have been in the past), you’re at least nominally familiar with ads on their platform. But as a reminder, on desktop, ads have two places they appear: in your News Feed, and on the right column. On mobile, they’re only in the News Feed. When ads show up for a Facebook page, that’s sometimes influenced by your friends’ interactions with that page — for example, if a friend has liked a new page recently and that page has set up sponsored ads, they’re more likely to show up on your News Feed.
There’s a few different types of ads for Facebook and Instagram that can be angled to pursue users at different stages of the sales conversion funnel, i.e. awareness, consideration, and conversion. These ad options include video, photo, slideshow, carousel, canvas, and dynamic product ads, which present themselves in the News Feed or sidebar (again, the sidebar only shows up for desktop users). Search ads are also an option but only appear for mobile users. Also, search ads are only possible when running concurrently with a News Feed ad campaign.
Instagram has a similar ad situation in that the primary placement for ads is in the feed. You also have another option for advertising: Instagram stories. While Facebook has (relatively recently) added stories to their platform, Instagram stories are far more central to users’ experience. Instagram story ads have the advantage of being super customizable, creative, and engaging thanks to Instagram’s built-in customizing options, although the downside is that users can click through them quickly if they don’t have a vivid enough visual hook.
About A/B Testing…
Now, on to the main point of this post: A/B testing, which is a way to compare two versions of an ad or web page to see which performs best. Essentially, it’s basic 6th grade science applied to online advertising; you take one ad that’s the control and another ad with a single concrete difference, then measure the resulting metrics.
Before starting an A/B test, it’s important to identify your KPIs (key performance indicators) through which you’ll measure which version is more “successful.” While you’ll probably end up looking at a few different ones, it’s also a good practice to identify one “primary” KPI. This is pretty intuitive, but it’s particularly important because different ads will have different purposes. So, if you’re attempting to raise awareness, for example, it will do little good to measure revenue. Despite the long term goal for, well, all your online presence being to raise revenue, the whole point of increasing awareness is to establish the foundations for greater revenue down the road. With that in mind, measuring revenue in this scenario instead of a KPI that relates to awareness could lead you to make the wrong decision in an A/B test that doesn’t actually increase the metrics necessary to increase sustained revenue in the later time frame in which revenue would be the appropriate KPI.
In other words: Identify a clear “what,” and a clear “how,” and you’re set up for success.
You also want to identify what variable will be changed in the non-control version of your ad. Just like with how your KPIs should have a clear reason for being what they are, you want to have a clear logic for what you’re changing. Sure, you could just throw different changes in there willy nilly and you will get data at the end of it. But, since ads cost money and time, it pays to be deliberate and precise. These test variables could be many things but some common ones are ad or call to action copy, imagery and other creative, or audiences being targeted. And again, back to 6th grade science: whichever variable you choose to change, everything else stays the same.
A/B Testing with Facebook’s Split Test Tools
The concrete step-by-step process of realizing an A/B test varies depending on the platform you’re using. Some website hosts have built-in A/B testing tools or, unsurprisingly, Google Ads has A/B testing capabilities. For your Facebook- and Instagram-based ads, however, Facebook has their own tools.
Facebook ensures, for example, that your sample audience is split into random groups that don’t overlap so that your results’ validity isn’t tainted by a non-controlled audience. They also measure each ad’s performance based on your designated KPIs and present that data to you internally. With the help of these tools, you can then implement the winning strategy across your whole campaign.
Digital ad campaigns can sometimes be overwhelming; there’s so many factors at play, plus the web is gigantic. When you’re wondering which direction to head and how to best make that happen, A/B testing helps break it all down into easily digestible parts. And with the presence Facebook and Instagram command in much of society’s lives, utilizing these platforms helps further pinpoint your advertising — while well-executed testing makes sure you’re getting the most out of them.
