Commentary
Twitter, TV's New Water Cooler
- by Dave Morgan , Featured Contributor, February 11, 2010
I've written before -- a bit tongue-in-cheek -- about whether Twitter data could ever replace Nielsen as a better way to measure television viewership. I'm going to revisit this issue again, since it's clear that Twitter data can measure something very important to media companies, marketers and their agencies, that Nielsen television panel ratings cannot. Twitter data is capturing the long-elusive "water cooler" chatter that so many media companies and marketers are trying to influence.
advertisement
advertisement
The chart below, taken from Twitter's blog, graphs the millions of Super Bowl-related updates the service received during the game as a percentage of all Twitter updates during that same time. The blue line represents those with words directly related to the game. The red line represents those with words related to the ads and advertisers. Yes! This data actually tracks immediate viewer reactions and "tweeting" related to the Super Bowl ads.
This is powerful stuff, particularly if you are Frito-Lay -- since the big spike in the red line (in section "C," on the left/middle of the chart) that eclipsed the blue line represents reactions to the first Doritos ad of the game. Twittering about the ad eclipsed Twittering about the game.
Does this mean that Twitter data is ready to ready to replace incumbent services in TV audience measurement? No. But it does mean that data from Twitter and other real-time web services are poised to become very critical components in the media and marketing measurement process. If one of the objectives of the Doritos ad placement was to get a lot of people engaged with it and communicating to others about it, Twitter's data proved that it accomplished its objective.
Arguably, data like Twitter's about actual audience engagement, reaction and sharing is actually more important that just ratings data. At the least, it is a great complement to it, particularly since it is based on hundreds of thousands of explicit and transparent consumer actions, not just projections from opaque panels. Watch out. I believe that we are just seeing the beginning of what will be an enormous, Web-service-fueled disruption in the measurement of media and marketing. Given his quote in the WSJ about Twitter becoming the new "water cooler" of TV, I wouldn't be surprised if Les Moonves didn't agree. What do you think?
That's incredibly fascinating data Dave.
It makes me wonder though: In a Super Bowl-like situation, where you literally have 100 million consumers all doing the same thing at the same time, what's the value of a Tweet sent by a user viewing Doritos' TV ad, to another user already viewing the same ad at the same time? It seems redundant and possibly counter-productive: Let's say Coke is running a :60 ad. The Tweets start flying at :15 or :20 into it and the next thing you know, nobody is paying any attention to Coke's $5,000,000 spot. Suddenly, the brand message Coke really wanted to get across is drowned-out in a tidal wave of back seat Creative Directors. Maybe they managed to extend engagement for another minute or two but is Coke net ahead in communicating their brand message? If not, future advertisers would be wise to run much shorter messages to A) Save money and B) Get their message in before the Twitter Wave hits.
On your other point Dave, I'm not sure that social networks per se will ever become viable replacements for media measurement panels...for a number of reasons. That being said, I'm absolutely convinced that a web-supported system is not far off...which will have a dramatic effect on the 160 or so markets still using paper diaries.
This is brilliant insight -- wow. thank you for sharing it so eloquently.
But what exactly does this mean? Words were mentioned, but in what context. And from there, what does it mean for the brand, what are the results?
For RGB or any first-timers, the mention is in itself a win. Not so for Frito-Lays, we all knew them already, what did these "mentions" get them and was it worthwhile.
Not degrading these finds, they are intriguing, but it still needs to be more than this to mean something.
Michael .. not sure exactly how to interpret mentions ... then again, the mentions (or recall) of brand names are what drives most post-campaign ad effectiveness studies. Certainly, each of the mentions could be looked at in context with the words surrounding it ... were the mentions positive? negative? related to brand attributes? competition? The point here is that Twitter data is becoming a very powerful source of information about how consumers react to commercial communication on television and the data has the potential to give a better view into the consumers' true state of mind that almost anything that we have ever seen before.
Very insightful and interesting, Dave. Thanks for sharing.
Statistically, it's significant enough to monitor. It measures buzz factor. But a quick look at the verbatims might yield less engagement than first thought. It's still one step ahead of Nielson, because you get real, unsolicited reactions in real time.