How Sentiment Analysis Helps Brands Sell - eMarketer

Newsletters Sign-Up

Plans & Pricing

Does My Company Subscribe?

How Sentiment Analysis Helps Brands Sell

June 29, 2016 | Marketing

Patrick Charlton
Director and Co-Founder
Buzz Radar

Sentiment analysis is already an important component of many brands’ social media strategies, but it can often be limited to basic interpretations of whether a conversation is positive, negative or neutral. At the Cannes Lions international advertising festival in June, data visualization technology provider Buzz Radar conducted an experiment that took sentiment analysis further, diving deeper into different types of emotional nuances. Patrick Charlton, director and co-founder of Buzz Radar, spoke to eMarketer’s Maria Minsker just before the festival about what the company hoped to learn from the project.

eMarketer: Before we talk about emotion specifically, can you provide an example of how brands can leverage real-time social conversation visualization technology in general?

Patrick Charlton: Burberry has used our Command Center platform to look at conversations on social media surrounding their campaigns. We pull in every single mention of Burberry from conversations about London Fashion Week, for example, and analyze the sentiment. They’ll have a big screen up during the 20 minutes of their live show, watch the conversation on a dashboard and choreograph the program using that live data. They can pick up the phone and talk to their content producers to optimize the show in real-time.

eMarketer: How are you incorporating emotion into this existing data visualization technology?

Charlton: During Cannes Lions this year, we are [combing through] every single mention of the hashtag #CannesLions in real time. That’s about 5,000 tweets per hour at its peak. We are running that through our platform as well as the IBM Personality Insights engine to determine the emotion and personality of the conversation every 15 minutes. We are visualizing it, and putting it up onto a giant Clear Channel billboard on top of a hotel in Cannes.

eMarketer: How does IBM Watson power the Personality Insights engine? How does the technology pinpoint the overall emotion of a conversation?

Charlton: There are two elements to it. The way it’s been described to me by the Watson team is that it’s an artificial intelligence psychologist. It’s been programmed to analyze lots of different types of conversations and then read the mood of those conversations to understand the tone and the subtleties. It’s not just looking at positive, negative and neutral, but really pulling specific emotions that people are feeling, [such as anticipation.]

“[IBM Watson is] an artificial intelligence psychologist ... programmed to analyze lots of different types of conversations.”

eMarketer: How will emotion analytics play into real-time marketing? What are some advertising takeaways that you hope to gain from this experiment?

Charlton: There are lots of different applications for it. The experiment at Cannes is about figuring out whether [brands can use] emotional responses. When you’re looking at an event such as Cannes Lions or the Super Bowl, what Clear Channel is interested in is whether or not it’s possible to sell emotions to different brands.

In other words, Nike might pay for an audience experiencing “excitement” or adidas might pay for an audience experiencing an “authority-challenging” emotion. Then, platforms can effectively trigger different advertisements depending on the emotion of the online conversation.

Go beyond the articles

eMarketer PRO customers get quick and easy access to the exact data and analysis they need to make critical business decisions:


  • Go beyond the articles:

    eMarketer Products

    You've never experienced research like this.

  • Hear from our clients:

    Customer Stories

    Nearly all Fortune 500 companies rely on us.

  • Want to learn more?

    Contact Us

    Inquire about corporate subscriptions today.