How Managers and Their Teams use Emoji Differently at Work
With the rise of digital chat platforms like Slack and Microsoft Teams, emoji have become a significant element of business communications. Emoji have a significant effect on communication via these digital channels: 81% of people think emoji users are friendlier and more approachable, 74% think they make positive news sound more sincere, and 65% of people feel more comfortable expressing emotion via emoji than talking on the phone.
For remote employees using digital communication in place of in-person conversations, emoji can help take the place of facial expressions, jokes, or even tone of voice. As the coronavirus pandemic forces many businesses to switch to a work from home model (either temporarily or permanently) these digital channels will only become more common.
Despite this growing role in business communications, emoji use hasn’t been rigorously studied; most existing research is based on public social media content or surveys. To help rectify this deficiency, our team used Cultivate’s AI/ML platform to anonymously analyze how leaders use emoji while communicating with their team members at work.
Our platform parses through digital conversations, first classifying messages by intent and affect, then determining how individuals tend to respond in various conversational situations. This gives us the unique ability to not only look at summary distributions of emoji usage, but to also dive deep and analyze the utility of emojis in directing conversation.
We found a variety of interesting trends including the top positive and negative emoji used by leaders, the differences in how managers use emoji versus their team members, how emoji use differs by company and more.
Data and Methodology
This analysis pulled anonymized emoji reactions from four different enterprises’ Slack messages over a 180-day period ending June 10, 2020. The total enterprise data set included 83,055 messages, 101,134 emojis used and 466 different types of emoji (custom Slack emoji were excluded). We also extracted metadata on the messages the emoji were responding to and the sentiment of the emoji themselves.
In order to establish emoji sentiment, we utilized a separate large public dataset from Twitter and Reddit to calculate a baseline for how positive or negative each emoji is. We determined the positivity of each emoji by averaging the sentiment of the messages in which they appeared.
One interesting finding to note: there was very little change in the data from Q1 2020 to Q2 2020, so the increase in remote work thanks to the coronavirus pandemic did not appear to have a significant impact on emoji usage.
Here are our other major findings.
Most Common Emoji Overall
The most commonly used emoji overall were, in order, Thumbs Up, at 30%, Heart at 8.14%, Laughing Face at 6.35%, Fire at 4.57%, and the Noisemaker at 4.37%.
As you can see, Thumbs Up is the most popular by a wide margin and all of the top five emoji had a positive sentiment. In contrast, negative emoji are very rare. The most popular negative emoji (Mask Face) made up just 0.27% of our data set. Frowning Face is second at 0.21%, Thumbs Down is third at 0.1%, Money Bag is fourth at 0.09% and Upset Face is fifth at 0.04%.
We also found that people tend to stick to the same set of emoji. 71.2% of users used fewer than ten emoji in the last 180 days and 50.7% of users used fewer than five emoji in that time. It’s not clear from our data if this is due to personal preference, or if the “Frequently Used” and “Handy Reactions” menus in Slack encourages people to stick to the same set (likely some of both).
Emoji Use Varies
Our data clearly showed that managers use different emoji than their team members. The top five emoji used by managers are entirely different than the top five used by their team members. In general, manager emoji tend to show positivity and appreciation (like the Clapping Hands and the Noisemaker) and team member emoji tend to show acknowledgement (like Checkmark and Eyes). We were pleased to see so many positive and encouraging emoji used regularly by managers!
Emoji usage appears to differ from company to company as well. We limited our analysis to the common set of emoji offered by Slack, reduced it to two dimensions with PCA, then fit a Gaussian distribution to the emoji usage of each company’s users to visualize differences in usage across companies (see the graphic below). As you can see there are distinct differences and overlapping areas. This could be caused by a variety of elements including company culture, differences in emoji vernaculars, team structures, etc. It seems that once a company establishes a commonly used set of emoji (which could be affected by their culture and the example that leader set), people keep using the same ones over and over again.
Getting Particular with Emojis
There are significant differences in emoji usage based on the intent of the message that the emoji are applied to. We calculated the global usage rate for each emoji, and also the usage rates of each emoji in response to messages of a specific intent (we’ve fine-tuned BERT to classify messages into a number of conversational intents, such as “giving praise”, “scheduling a meeting”, “requesting information”, etc.). The differences tell us if people are more likely to use certain emoji in some conversational situations, and if so, which ones. The values below are the differences between situational usage rates and global usage rates.
For example, an ad-hoc scheduling request will receive the Thumbs Up emoji 16.3% more often than general posts will. Posts that give recognition receive Clapping Hands 6.11% more and the Noisemaker 4.34% more. Posts that inform that work has been completed get Thumbs Up 14.64% more and posts that express doubt also get the Thumbs Up 9% more. While it’s clear that the Thumbs Up is a very popular emoji overall, less-common ones also appear more often in specific niches. For example, ad-hoc scheduling requests are also more likely to get the Coffee Cup emoji and messages informing that work has been completed are more likely to receive the Rocket Ship.
Scheduling requests and making commitments elicits the highest sentiment emoji responses (average sentiment of emoji responses, per message instance), while expressing disagreement elicits the lowest sentiment responses. However, it might be unwise to read too much into “emoji sentiment” – the low sentiment emoji responses to disagreements could be showing support or alignment with the disagreement (for example, if someone posts a message saying “I don’t think this is right approach” then a Thumbs Down emoji might mean the poster agrees that the approach is wrong.
All in all, emoji use is somewhat customized based on an individual’s company and position. Overall, it’s extremely positive, with positive sentiment emoji used far more often than negative ones. Next time you write a Slack message consider what emoji you’re using and why – it just might help improve your digital communications with your team.
About the author: Andy Horng is the co-Founder and Head of AI at Cultivate, a platform helping companies leverage their wealth of digital communications data to strengthen workplace relationships and unleash leadership potential. Prior to founding Cultivate, Andy worked as a machine learning practitioner across a variety of domains, including tracking dementia via neuroimaging, developing product recommendation engines, and automating information extraction of legal contracts.