“Bluefin Labs is the first company to start at scale linking publicly available conversations to TV viewing data,” says Deb Roy, co-founder and CEO of Bluefin and director of the Cognitive Machines Group at the MIT Media Lab. “We’re turning this data into a feedback loop that allows marketers to move beyond focus groups and surveys to get real-time information about people’s reactions to their products.”
The ability to tie together traditional viewing data such as Nielsen gathers with moment-to-moment ebbs and flows in conversations about any given brand, event or program aired on TV represents a fundamental shift in marketing techniques where cognitive science and psychographic become basic tools of the trade. “We can curate the chatter now radiating through layers and layers of social groups with viewing data to show how impressions are driving expressions,” Roy says.
Bluefin is a Web-based platform that tracks over three billion social media comments per month to provide comprehensive analysis of viewers’ responses to over 210,000 individual TV telecasts airing across more than 200 TV networks. “We believe we have the most complete index of TV anywhere,” Roy says, explaining that the index is built from metadata, closed-caption text, fingerprints and other information associated with each show.
Viewing and social commentary are tracked with time-shifted as well as live programming. In fact, Roy adds, “70 percent of our data link to time-shifted viewing.”
All of the TV and social media data are aggregated into a massive database Bluefin calls the “TV Genome,” which it defines as “data created by the mapping of social-media commentary back to its stimulus on TV.” A dashboard graphic rendering of the genome provides a constantly shifting snapshot of real-time chatter across millions of connections, allowing users to sort through by program category, time of day or any other dimension and then drill down from the mass aggregation level to very specific clusters of comments around specific show segments.
The firm’s flagship service, Bluefin Signals, delivers a wide range of graphics tools that render the analysis of social media data and viewing data into charts that help marketers better understand what’s driving the appeal of their brands and programs. For example, features known as Audience Profile Layer and On-Demand Audience Profile Creation give Bluefin clients the ability to analyze the television preferences of specific audience segments.
“You get new views of what’s happening with audiences,” Roy says. “You can see patterns never seen before.”
Bluefin has signed up wide range of clients since its commercial launch in July, including A+E Networks, AMC Networks, CBS, Comcast Spotlight, mcgarrybowen, MediaCom, Media Storm, Pepsi and Turner Broadcasting System, Inc. “For TV networks, being able to mine social media commentary about TV represents great promise, but the data needs to be precise and the insights need to be specific,” notes Jack Wakshlag, chief research officer at TBS. “Bluefin Labs’ analytics platform plays an important role in supporting these efforts.”
The use of cognitive science and advanced quantitative analytics has become an important new trend among ad agencies, which are adding experts in these fields to their staffs and, in some cases, have acquired firms specializing in these capabilities. “Prolific social media use gives us incredible insight into customers’ preferences and behavior beyond just demographics,” says Anush Prabhu, managing director of communications planning and analytics for the mcgarrybowen agency.
“Our clients tend to spend the bulk of their media monies in television,” Prabhu says. “Bluefin Labs’ data provides the ability for mcgarrybowen to harness social media insights towards the planning and optimizing of television media and creative, ultimately allowing us to deliver more integrated and efficient campaigns for our clients.”
Bluefin has begun exposing some of its research findings on its Web site to help people understand what the platform delivers. The free Signals Pulse dashboard provides “overnight ratings” from social TV chatter on top-rated programs that were viewed the day before and drills down into how programs or ads track with certain user categories, like hard gamers. Tracking goes much deeper to correlate various groups’ responses with segments within shows, even to the point of categorizing real-time comments on a second-by-second basis.
As an example, Roy shows how campaign managers could see how their candidates’ comments in debates are playing out in the social sphere on any given topic. “Just measuring the volume of expression is not enough,” Roy notes. “We can drill into available data using psychographics to get deeper insight into audience behavior.”
A marketer might zero in on people who are avid consumers of a particular soft drink by identifying who tends to send comments out on Twitter when an ad for the drink appears. This information can be used to determine what shows these people are most drawn to, which in turn can be used to better discern which shows are most appealing not only to current consumers of the drink but potential consumers as well.
For example, in a recent demonstration of the platform Bluefin identified consumers of Mountain Dew and then tracked what shows drew the most social media commentary from these “Dewheads.” It turns out these people tend to be enthusiastic about comedy shows, often registering high social rankings for programs which have middling popularity with the general population.
Given the irreverent comedic theme at the core of the brand, the demonstration offers a strong affirmation of how tastes and attitudes can be tapped through smart branding and how the effectiveness of that branding and placements of advertising are tracking with marketers’ goals. “How you advertise – the branding, creatives, placements – it’s all up for redefinition,” Roy says.