« Back to Posts

30000WAYS TO SAY I LOVE YOU_TVREV

30,000 Ways To Say I Love You

Love is complicated. So complicated, that when it comes to social TV, there are over 30,000 ways to say it.

Just check Twitter Wednesday nights at 8, and you’ll be bombarded with reactions like Alicia gives me so much life. #Empire and Yassss #Empire. It’s clear these fans are emotionally engaged. They are leaning in, and commenting on what’s happening on screen in real-time.

In fact, Nielsen recently released a study concluding that when fans tweet live, 65% of them are tweeting about what’s happening on screen.

But do they actually like what they see? As the content paradigm continues to shift from live experiences to on-demand ones, there is a growing insight deficit on how to effectively connect emotionally with fans. The marketing and advertising industry has been stuck reading lukewarm green, red and gray sentiment charts for the last five years, trying to decipher if content and ads resonate emotionally.

Love, Meet TV.

The first wave of social TV taught us that the purpose of second screen apps, whether measurement or engagement platforms, should be to drive live viewership or produce revenue. As the always-on content consumption model is more widely embraced by content creators, distributors and advertisers, the third purpose of second screen apps will be to drive emotional investment in content.

The problem, of course, is that the industry’s options in measuring emotional investment in content are limited to traditional market research or social media sentiment analysis. In 2014 there was $4 billion spent on traditional market research in the media space, including focus groups, opinion polls and surveys. These methods have small samples, and they’re laborious, expensive and not real-time.

Then there is social media sentiment analysis (cue eye roll from marketers). Most automated sentiment analyses represent arbitrary assignments of “positive”, “negative” and “neutral” to social conversations, which shove complicated human language into unrealistic oversimplifications. This approach is especially uninspiring given the unique nature of TV series, where villains and plot line twists are meant to be hated or crazy. And that’s a good thing right? Ipso facto, nobody trusts the current crop of sentiment analysis providers.

Trust demands symmetry of data – understanding the information in the context of what you already know. Social always feels splintered and unorganized because it’s rarely presented in the context of solving marketers and advertisers’ problem. To make it worse, language has become nearly impossible for computers to understand, a point punctuated by our recent analysis of a millennial focused TV show, where 67% of reactions were completely unrecognizable by state-of-the-art sentiment analysis.

The Business of Emotion

Emotions are a common language that content creators and advertisers have been eager to understand since the advent of market research in the 1920’s. As Twitter’s COO Adam Bain recently said, “[t]he entire monetization business is the business of monetizing emotions.”

With that in mind, the ecosystem is evolving. Creators are looking to analyze the events which precipitate shifts in the audience’s emotion so they can adjust storylines. Marketers are yearning to understand why their content resonates emotionally in order to maximize spend and better facilitate ongoing dialogue with fans. Most importantly, advertisers have a voracious appetite to invest in content that fans love (“follow the attention” as they say) as it ensures the audience is paying attention.

So, if the final purpose of Social TV is to drive emotional investment in content, how do we turn the billions of social TV conversations into insight into what audience’s love?

The Modern Theory of Emotional TV

The answer is right under our noses when you consider that fans expressed how they felt about live TV 220M times so far this year alone on Twitter.

At Canvs we measure emotional reaction to content, integrations and advertising using social data. We define a Reaction as a social conversation with at least one emotion present. Each Reaction can fall into 56 different emotional categories, and a multiple more due to Reactions with conflicting emotions.

Emotion isn’t just interesting from a qualitative standpoint. When measured correctly, it drives meaningful outcome. The prevalence and proportion of our Reaction data has been successfully correlated to show loyalty, series renewal, and next episode ratings. The potential correlations are endless.

As an industry, we implicitly know how important driving emotional investment in content is, but don’t know how to accurately measure the emotions our audiences express every day. Social media, analyzed well, is a true lens into how people feel and represents the crucial convergence of social measurement and social science.

In a world where expression is democratized and everybody has a voice, we need to advance the way we understand audiences. And to start feeling the love, no matter how it’s expressed.