Amazing technologies that measure, in a deep way, how people respond to marketing messages were presented at the recent Media Insights and Engagement Conference. But it’s safe to say that despite great potential, some efforts do not scale well in terms of cost, logistics or time-to-insight.
And. of course we have the old standby, focus groups, for whose powerful insights businesses spend an estimated $40 billion a year. Unfortunately, like some of those technologies at the conference, focus groups also can be expensive, hard to scale and time-consuming.
All these technologies are trying to do one thing: measure consumers’ emotional responses to a message. Out of emotion comes intentions. Out of intentions come behaviors. Thus, emotions can drive business outcomes and the bottom line.
Emotions matter.
But accessing and understanding the emotions people have can be, yes, expensive and time-consuming. Thankfully, we have a powerful new “focus group” sitting in front of us that promises to be an agile, affordable new tool for marketers to understand how their messages are received: Twitter.
advertisement
advertisement
Every year, Twitter generates about 1 billion opinions that express how people feel about every corner of television, from the shows to the commercials to everything between.
Yes, companies routinely monitor social media (spending an estimated $200 million annually to do so), but that monitoring is usually colorless or too focused on measures such as volume rather than on the emotions people routinely express in those many millions of tweets.
At best, many of these systems use a simple sentiment analysis to assess whether a post was positive or negative. That provides some feedback, but falls far short of capturing the rainbow of expression that Twitter routinely carries.
Beyond that, the nuances of fast-evolving online language typically elude traditional sentiment analysis. When a tweet calls a show “sick,” is that positive or negative? Is it a fusty moralist harrumphing, or a skater bro enthusing? Natural Language Processing was invented decades ago. It is not “on fleek.”
Finally, we have the technological horsepower to evaluate how Twitter (and soon, Facebook) users are talking about TV shows and commercials, working with data parsed initially by Nielsen.
This approach holds great promise, particularly as we gain deeper understanding of the emotions expressed on social media, and as a series of trends make this kind of broad data more valuable than ever:
This creates the world’s biggest (and most relaxed and most honest) focus group. It’s sitting in front of us, providing more feedback every day. And finally, we have the technology and processing horsepower to actually understand, in close to real time and reasonable cost, what people say they’re feeling about a show, a brand or a message. Even better, it’s largely unfiltered, talking in the moment about TV and the brands that appear there.
We can already predict, three weeks before an announcement, with about 85% accuracy, which TV shows will get renewed, based on the emotional engagement and response those shows can drive online. Deep understanding like that will only grow, and can be powerful for both brands and creators.
We want to understand what the world is thinking, in a fast, cost-effective and meaningful way. The industry continues to devise new ways to harness different technologies to do that.
But social media provides another source of fast, affordable and broad insights. Millions of people in the world’s largest focus group are telling us how they feel. We can make magic with that.
Jared...an interesting piece. There are advantages to both social data and traditional focus groups as we know. Certain stuff simply can't be probed via online to the extent offered by a moderated face-to-face situation. Using the two in tandem, social data provides some excellent discovery opportunities for traditional group agendas. One question though; how does one separate the more cavalier emotional material often found in social from the deeper emotions that 'might' drive behavioral intentions?