« Back to Posts


Google For Emotions Promises Empathy, And Hit Shows

Artificial intelligence (“AI” as we largely accept it) manifests in the world in countless ways, and has varying definitions. Some fear AI because of its ability to perform tasks that normally require human intelligence. But what about AI that offers something humans need help with — empathy.

People so often misunderstand each other. We hear or see words and we interpret them different ways than the speaker or author intended. In one sense, modern language itself has become a hieroglyphic tapestry of phrases, jargon, misspellings, hashtags, and emojis that may not necessarily hold the same meaning to the author as it does for their audience.

Enter Canvs. TV[R]EV has documented Canvs’ rise as the leader in emotion measurement, working with TV networks, digital publishers, streaming platforms and brands alike in order to analyze the emotional underpinnings behind the words people use online to express their feelings about content. The company’s Emotional Reactions (ERs) have become somewhat of a trading currency in media, where traditional and digital media can get a holistic sense of character or plotline resonance, audience engagement, show affinity and more.

But rather than simply analyzing social comments surrounding existing programming, Canvs now has its eyes set on the multi-billion dollar research and consumer survey industry.

To that end, the NY-headquartered company announced today via TechCrunch that it’s rolling out Canvs Surveys, an emotion measurement AI engine that analyzes open-ended responses. Canvs Surveys ingests spreadsheets filled with open-ended verbatims and — in minutes — spits out a comprehensive, compartmentalized analysis of the qualitative emotions respondents’ expressed in response to survey questions. 

In doing this, Canvs’ emotion measurement insights can be used in more prescriptive, business-critical ways. For example, explained Benoit Landry, senior director of program research at NBCUniversal, “Leading up to upfronts is this TV hunger games. We’re talking about a four-to-six week period for every broadcast network, and the challenge is to try to decipher which of these pilots—and there can be as many as 30 pilots—have the most potential for success as a series. And that’s been a very difficult task.”

NBCU is judging its pilots and picking “winners” based on the Canvs Surveys analysis of open-ended responses. And NBCU went from spending 16 hours trying to hand sort open-ended survey responses, down to one hour with Canvs Surveys. Those saved resources scaled over time should offer monumental savings and afford plenty of new opportunity growth.

It’s also worth noting that Canvs has its eyes on the broader research market beyond its media and entertainment bread and butter. The company also announced today it added Kristin Luck to its Canvs Board of Advisors. Luck is a pioneer in marketing measurement. The growth-consulting expert and proven entrepreneur co-founded the pioneering online research firm Online Testing Exchange. OTX, now part of Ipsos, was named the fastest growing research firm in the world in 2002 and 2003. Luck, who now runs her own consulting practice, also was CMO and president of Decipher, a marketing research services provider, after it acquired her startup, Forefront Consulting Group.

Kristin offers a unique vantage point into how our core language-emotion AI can be leveraged across industries beyond media — anywhere research and open-ended surveys are conducted,” said Jared Feldman, CEO at Canvs.

And the broader focus is a financial-driven mission worth having. Recent forecasts predict worldwide revenue from emotion analysis software will increase from $123 million in 2017 to $3.8 billion by 2025. One can imagine the leading emotion measurement AI has an excellent shot at addressing much of the need in this burgeoning market.

Also, seeing is believing, and Canvs is giving any researcher the ability to see Canvs Surveys in action at no cost. Beginning today, Canvs Surveys is available for any researcher, where Canvs will provide the analysis of the first 5,000 open-ends for free.

Feldman closes the discussion well with his broadest intention: “Our goal is to create a Google for emotion.”