Introducing Canvs

At Mashwork we tell stories with data. This is the story of how we used data to build a social TV insights platform unlike anything you’ve ever seen.  Meet Canvs. I learned early on that data is only as good as the story it tells. When I was in college, I had the chance to reverse mentor several executives at major media companies. They didn’t want to know about all the tools flooding the social media analytics market. They just wanted the answers to their questions—things like, “How did our TV show do last night?” and “Are we reaching our target demographic?” In trying to answer their questions I tried nearly every tool on the market and realized that there was a huge gap between the questions people had and the tools available to them.

Canvs started by addressing a problem—none of the tools on the market were capable of providing insight and context instead of raw, complicated data. I joined forces with Professor Sam Hui, PhD, a brilliant data scientist I met at NYU. We soon became completely immersed in the market, allowing us to gain a fine-tuned understanding of the problems analysts face. We did this by becoming analysts ourselves.

Next generation social TV insights had a humble beginning

We’re proud of how long it took to build Canvs

I founded Mashwork in 2010 and patented the technology powering Canvs in 2012, but it’s taken us until now to launch. To build an accurate product you need a lot of human input and domain knowledge. While our competitors splintered off and built all sorts of tools and service offerings, we focused like a laser on delivering insights for entertainment clients. Everything we did informed the development of Canvs.

Canvs Social TV analytics in an early incarnation

We don’t care about sentiment

Once you see Canvs, you’ll know it’s different from every other platform you’ve used. All the unsupervised sentiment tools out there spit out positive, negative, or neutral classifications about posts. We know those classifications aren’t actionable or accurate. We don’t care about sentiment. Instead we care about delivering a glimpse of the nuanced range of emotion your audience expresses. You won’t see the words  “positive” “negative” or “neutral” anywhere in Canvs. Instead, you’ll see words that reflect the way real people talk like “LOL,” “awesome,” and “shocked.” And Canvs doesn’t care how you spell things or what slang you use. We continually teach the product to recognize all the ways people talk about entertainment content.

Social TV insights from Canvs

Canvs is beautiful

Part of what makes Canvs empowering is that we put design first. Most product designers make the mistake of thinking that B2C products have to be beautiful while enterprise products simply have to be powerful —they forget that analysts appreciate good design as much as anyone. We’ve worked with a lot of robust tools that are terribly designed, and we’ve also seen tools that are gorgeous but unusable. There’s a striking balance in Canvs’s design that allows you to synthesize the information easily. Canvs is the first system that organizes emotions in a way that’s immediately relatable to sophisticated analysts and to their bosses who haven’t ever touched any social media tools.

We can’t wait for you to try Canvs. Our current partners have been awesome in helping us build a product that’s as useful to them as it is to us. In the eight months we’ve been in private beta, we pushed more than 75 features, many based on suggestion from our partners. We think we’ve nailed giving marketers actionable social data for airings of TV shows. We’re improving Canvs every day to give you even more context about your content and audience. We want to help you answer your pressing questions as fast as we can. Come see for yourself at

Yours truly and forever hustling,

Jared Feldman

CEO & Founder, Mashwork

Canvs - The next generation in social TV analytics from Mashwork on Vimeo.