With Nielsen Partnership, Canvs Becomes First Fully Automated Social TV Sentiment Platform

By Adam Flomenbaum 

canvsCanvs has been providing us with sentiment analysis of tweets around TV shows for a few months now, but has today announced a formal partnership with Nielsen and its Twitter TV Ratings. Canvs, then, becomes the first completely syndicated qualitative social TV platform, giving both networks and marketers more tools to extract meaning from viewer tweets.

In the social analytics world, sentiment analysis – once hot – has been largely cast aside. We asked Mashwork Founder & CEO Jared Feldman about this and he said that sentiment analysis “has been dismissed out of hand for so many years” because the solutions have been incomplete.

“We built Canvs to be actionable because the “positive,” “negative,” and “neutral” categorizations that come standard with most social media tools simply aren’t useful to marketers,” Feldman said. “Automated tools are unable to parse complex emotions like “I hate how much I love Scandal.” Further, what if hate is a goal? While a Tweet about despising Joffrey from Game of Thrones means the production team at HBO did its job, a traditional tool would simply label that tweet “negative.”  Canvs not only tells you what was “good” and “bad” about a show, but it also tells you which characters were attractive or annoying and which plot lines made people say WTF or SMH. These nuanced insights give marketers the ammunition they need to react to how their audiences really feel.”


Feldman acknowledged that there other tools on the market that perform nuanced sentiment very well, but these tools are time and labor intensive. In the social TV space – where relevance has a deadline – Canvs will give networks and marketers immediate results.

Immediate access to this data is compelling, but only in the context of how it can be leveraged.

“We find most of our clients are using the social TV insights from Canvs to inform broad marketing and programming decisions rather than only using Canvs to inform social executions,” Feldman told us. “We’ve had clients use Canvs to identify the best clips to use in promos on-air and online based on how audiences react to certain plots or characters. Canvs can also help reality TV producers determine which characters or contestants to feature in future episodes. We think of Canvs as a real time opinion poll—it’s like having a focus group of your most passionate fans volunteering their feelings about TV content.”

Production powerhouse Endemol Beyond USA is already leveraging the platform. “Canvs has proven itself to be a great tool for learning more about our audience and their opinions than other tools, Jacob Shwirtz, Chief Social Media Officer, Endemol Beyond USA, told Lost Remote. “This has broad applicability and we’re actively experimenting with how Canvs data can inform decisions in marketing, programming, casting, editing and more.”

c3 c2