Webinar Recap: Making the Most of Sentiment and Emoji AnalyticsApril 14, 2016
Emojis are more than just the “slang” of modern-day English – it’s a language that spans the globe and provides multidimensional insight into the emotions that people are feeling.
Melissa Fares, social media reporter for Thomson Reuters, is using emojis and sentiment analytics to enrich her coverage of breaking news.
Meanwhile, our Senior Analyst Alan Steel is studying if social media is better at predicting presidential election outcomes than traditional polling.
Both Melissa and Alan joined us this Tuesday to discuss the power of emoji and sentiment analytics for journalistic and marketing applications. Below top highlights from the conversation.
Elections and Social Media
“This year has been a game changer for use of social media, especially by candidates.”
Since August 2015, Zoomph has collected over 175 million social conversations around the U.S. presidential candidates.
With the youngest millennial now being 19, the demographics of modern-day voters has shifted to include digital natives. As a result, social media has evolved into a key vehicle for candidates to reach and understand their full voter base.
Over the last several months, we’ve seen some candidates rise up to the occasion and take the social game to another level. But others failed to gain momentum in the running for America’s Next President, simply by remaining out of touch on social media.
Benchmarking to Spot Sentimental Trends
As illuminated by Alan, gathering the right measurements on social media “emojins”(emotions as told by emojis) and sentiment require careful setup of your tracking engine.
To glean meaningful insights from your data over time, you must establish a benchmark average prescribing norms, since seldom will any person or brand receive 100% of love all the time.
The trick here is to find the ratio of love vs. hate that a person normally receives so that you can better spot an abnormal blip in the charts.
Alan applied this consistent tagging and benchmarking methodology to his study on the 2016 presidential race. He has been comparing each candidate’s ongoing performance to their share of positive sentiment online.
— Alan Steel (@alanwsteel) February 27, 2016
By doing so, he was able to identify Marco Rubio as the winner of the Feb. 26 GOP debate purely by looking at social analytics and seeing a dramatic rise of positive sentiment above Rubio’s normal benchmark score.
Today he reminds that any brand can choose to track campaigns, products, or competitors instead of candidates. They can asses their short-term and long-term successes with descriptive sentiment analytics.
Sentiment + Emojis = A Match Made in Heaven
Being the modern-day hieroglyphics that everyone and their mom uses, emojis provide context for the red and green that we see on our sentiment charts.
“The thing about sentiment is that it’s really a binary metric, meaning that you can see positive and negative sentiment. I look at it as a quantitative measure…but [emojis are] a qualitative measure.”
Zoomph’s Emoji Analytics make it especially easy to track the top emojis associated with any brand. It’s the first and only tool that can track emojis organically, without necessarily relying on hashtags or keywords for discovery.
While charting the top emojis associated with each presidential candidate, Alan predicted Jeb Bush’s fallout from the presidential race; the baby bottle and laughing emojis were clear indications of the public’s dismissal of Bush as a serious contender.
Alan was also able to spot a trail of breadcrumbs left behind by Trump supporters who trolled other candidates online. Notice the special appearance of Trump Trains in nearly every other candidate’s “top emoji” column…
Sentiment Analytics in News Reporting
Social media is green with news and opinions from all over the world. Melissa not only uses social media to locate trending topics, but to define what’s worth writing about.
While attempting to augment Reuters’ coverage of the U.S. elections, Melissa Fares used Zoomph’s analytics tools to gauge audience reactions and pinpoint the catalyst behind certain trends.
— Allison Lee (@ayylee07) April 12, 2016
She could measure the global impact that candidates were making in real time by harnessing the power of sentiment analytics.
As an example, when news first broke out about Trump’s Muslim-related comments, Melissa grappled the firehose of social activity and emerged with measurable insights. She reported that outside of the U.S., there were roughly 4.2 negative social mentions for every positive one regarding Trump. Inside the U.S., there were about 3.2 negative mentions per every positive mention.
Read full article: Trump’s call to ban Muslims draws fire on social media
The overarching lesson? We don’t need mood rings anymore to tell us how others are really feeling.
In another instance, Melissa measured the velocity of negative tweets tagged #NeverTrump. She was monitoring this when #AlwaysTrump cropped up on her radar.
“You had both of these hashtags ‘hashing it out’ on Twitter, both going viral…It was a war. People were going to war on Twitter.”
Using social analytics, Melissa could provide a holistic view of the tug of war between negative and positive sentiment online. When asked for a report, she was able to declare #NeverTump the likely winner of the trending war, noting how it was earning twice the amount of traffic as #AlwaysTrump.
Read full article: Trump naysayers push #NeverTrump on Twitter before Super Tuesday
Even now, Melissa is able to flesh out a number of multidimensional, data-enriched stories on headlining news. At a time when emotions are hot and social media is becoming noisy, she can rely on analytics to help tell her story.