Facebook Reactions are Useless for Brands

Facebook reactions are great. You can  a friend’s memorial post about their grandmother,  a post about pollution, and  a GoPro stunt video. Sadness for someone’s loss, anger at environmental neglect, and awe at daring feats.

But you could just as easily  the memorial post,  the pollution post, and  at the video. Empathy for the poster’s loss, shock at the state of things, and laughing at the insanity of a stunt. And then there’s always the old fashioned  , a standard acknowledgement of a post, and not necessarily a specific indication of feelings on the matter. In other words, similar actual reactions to a post can be expressed as multiple Facebook reactions. This is nothing new, the range of human emotion is vast and complicated, we get it, that’s why Likes evolved into Reactions. But what does this mean for businesses?

Internet chatter first hinted that Facebook’s Reactions are great for businesses, boosting engagement and allowing brands to know what their customers are “really” thinking. In reality, is that the case?

For starters, reactions have not resulted in higher engagement rates, at least not yet. And while the six reactions are perhaps intended to be face-value, this is not always the case (as I clearly demonstrated earlier). Because of this, Facebook Reactions will not give your brand an accurate depiction of customer sentiment

Let’s say that you post about a special offer you’re running on the brand’s Facebook page and you see some Angry reactions. They could mean “I hate this offer it is terrible,” “This is a great deal but I’m mad that I’ll miss out on it,” or “I hate this product!” Things really get complicated when you start looking at sarcasm. A Wow reaction could easily mean “wow that’s great!” or “Wow that’s soooooo great,” and those are two wildly different reactions.

   vs  

On the other hand, be wary if you’re thinking of pursuing automated sentiment. I ran a sentiment analysis on a sample of posts marked with “#sarcasm” or “#not” to see how the analysis labeled them. What I found was not comforting. Here are some examples of posts marked as positive:

“I remember how lovely the #English weather is #not”

“I love all these new @starwars fans. #sarcasm”

“Airports have the “nicest” carpet!!! #sarcasm”

“You know what’s nice? When collaborators decide they know better than you and ignore the carefully worked out protocol you sent them. #not”

“I love it when people laugh, as my toddler is kicking and screaming on the floor in the store #sarcasm #toddlerproblems”

“Omg I am still LIVID about this place charging me $180 to repair my broken windshield wiper. They’re getting a nice review on Yelp. #not”

To most human readers, these are obviously negative comments. But with automated sentiment programs, it is far more likely that negative statements are misinterpreted as positive, as opposed to the other way around. That makes sense – sarcasm is defined as the use of irony to mock or convey contempt – it’s not meant to be used as a positive reflection.

This brings us back to Facebook reactions. On it’s own, saying your post generated 172 Likes, 3 Loves, 15 Angers, 29 Wows, 14 Sads and 2 Hahas is effectively useless. What does that actually mean? What are your customers actually saying?

Facebook Reactions are a start, but there needs to be further, human analysis of your customers’ reactions. All in all, if you want to know what your audience is really thinking, it’s going to take more than a handful of emojis to tell you.

Angie Goldman
Marketing Analyst, Marketing Technology

[cta]