Do emojis improve customer service interactions? Let’s find out!

TwitterFacebook

The following is the first in a new monthly content series, called CX Data Lab, where we leverage analytics and data science from within the Simplr platform to uncover best practices and trends in customer service interactions.

🙂

Did you know that some historians place the first instance of an emoticon–or emoji–in the 17th century? In 1648, poet Robert Herrick wrote: 

“Tumble me down, and I will sit Upon my ruins, (smiling yet:).”

Pretty cool huh?

Now, other scholars argue that it was a mere typo, that both grammar and printing standards were too fluid to guarantee that Herrick’s then-quirky use of punctuation would be read and appreciated by others. 

Me? I choose to believe that the poet who famously revived the notion of carpe diem also preceded the more often cited first use of an emoticon by 335 years, when in 1982 Carnegie Mellon computer scientist Scott Fahlman wrote the following in an internal memo:

I propose the following character sequence for joke markers:

: – )

Read it sideways.

Whether you choose to believe the emoji has been around for 40 or 400 years, there is little doubt about its central place in linguistics and culture today. The emoji, in addition to GIFs, memes, and screenshots, has reconfigured the way we communicate with one another to rely less on words and more on images. We are, in a sense, living in a hieroglyphic renaissance.

But is it a better, more empathetic and personal communications tool than just text? Do emojis actually help to develop rapport between, say, a customer and a customer service representative? 

It was something that Ilias Miraoui, who also works on machine learning and AI here at Simplr, and I became curious about after noticing a small but significant bump in CSAT scores between customer service interactions that included emojis compared to those that didn’t. Specifically, we saw that interactions that included emojis had an average CSAT score of 4.7/5. All other interactions had an average CSAT of 4.6/5.

So, in the spirit of our data science team, curiosity won the day, and Ilias designed an experiment: he randomly inserted recommended responses for our specialists to use that included emojis to half of the conversations with customers, and in the other half of the conversations we kept emojis out. We did this to best control for any other factors that might influence the resulting customer sentiment so that the interactions were roughly “similar” in every other way.

In examining the data set, we found that 25% of tickets had a meaningful improvement in customer sentiment when an emoji was used compared to those in which no emoji was used.* 

🤯

That’s quite a significant impact by the small but mighty emoji!

(Now, I imagine some are questioning the impact of a .1 raise in CSAT scores. But even marginal improvements in customer experience–especially when the floor is already pretty high as in the case with a 4.6/5 CSAT score–can make a huge difference in loyalty and customer lifetime value. In fact, Harvard Business Review found that when customer experience scores went from a 9/10 to 10/10, the amount of revenue per customer went up 26%.)

So, should you stop everything you’re doing, get your representatives off the floor for an emergency best practices training session on emojis? If your CSAT scores are not where you want them to be, emojis are probably not the silver bullet. However, if you’re looking for some finishing touches to further take your CSAT scores to the next level, encouraging your representatives to throw a 😀or 😂 into a chat might actually do a lot to further develop personal connections between your brand and your customers. 

And that, as they say, is 🔥.

(Editor’s note: There is also a popular theory that Abraham Lincoln invented the emoji. As if that guy didn’t already accomplish enough.)

*For the data-inclined, you might be wondering how we define “meaningful.” We use a proprietary sentiment score that allows us to accurately gauge customer affinity in interactions with our specialists. That score is measured on a scale of 0 to 1. Anything that improves the score by greater than .2 is typically large enough to be considered meaningful, in our opinion.