Everything You Need To Know About A/B Testing Facebook Content (Give or Take)
by Adam Skikne
Social@Ogilvy recently conducted an international study that identified a significant decrease in organic reach for brand pages on Facebook. They found that organic reach for brand pages dropped to about 6% of fans, declining by an average of 49% during the three month study. For pages with more than 500,000 likes, the effect was even worse.
Now, if you’re a brand that has invested in building a presence on Facebook or if you work in social media, producing content for Facebook, you should find the results of this study quite worrying. Facebook wants to charge you to reach your own fans, but even when you pay them for the privilege of doing so, how do you get the most bang for your buck?
The answer could be A/B testing your social media content and adapting your paid media strategy.
Let’s start by looking at how the majority of brands go about creating and posting content on Facebook. Most brands post twice a day, once in the morning and once in the afternoon. These brands generally post at “peak times” (essentially when most of the other brands are posting), which probably means that this content is less likely to show up in someone’s newsfeed. In addition to this, most of this content is produced for the sake of producing content and filling up retainer hours. It’s not an ideal setup, but it looks something like this:
An alternative would be to incorporate A/B Testing into the planning, posting and promotion of your social media content. To do this, simply create two versions of the post you want to test. You can look at changing things like the tone of the copy, the call to action, the overall design of your image or even pit two completely different messages against each other.
Share both posts with your fans – both posts will most likely only reach a small percentage of your fanbase. Give the posts an hour or two and see if one post in particular has been more successful. To get the necessary exposure, you may have to put a small bit of paid media behind each post. In most cases, success will be higher levels of engagement (more likes, comments or shares) but this will depend on your objectives for your content. If there is a clear winner, promote that post to the rest of your fanbase or to a specific group based on their interests.
This approach looks something like:
If you want to test the same message with different audiences, post the same message twice but use targeted Facebook advertising to reach different audiences based on age, gender, location or interests, and then compare the results.
Now obviously, the degree that you incorporate A/B testing into the development of your social media content will differ from brand to brand. In an extreme case, a very data-driven client may want every single post to be A/B tested but this is most likely pretty rare. More likely, most brands will usually be open to use A/B testing when:
- Pre-testing a new designs for social
- Comparing results with different target audiences (supported by targeted Facebook advertising)
- Testing what call to actions drive engagement
- Compare click through rates from social content on your website (this may require tracking different links on each post)
- Optimising paid media spend
Eventually, you might be able to identify key trends that will allow you to target and engage different segments of your community of Facebook fans more effectively. This will allow you to ensure that the right messages are reaching the right audiences as well as make smarter use of paid media spend.
As I mentioned earlier, our job as social media people is not just to produce content for content’s sake and fill up retainer hours. If you are producing content for a client, explore and experiment with A/B testing in order make sure you are being as smart and strategic as possible when it comes to planning and creating content.