LexaBlog: Our Sentiment about Text Analytics and Social Media
Traditional vs. non-traditional media coverage
At Lexalytics, we host a demo site called PoliticalTrends.info that tracks and analyzes content from over 300 political blogs. We use it mainly as a forum to showcase our software and to have a “live” version of our capabilities up and running at all times.
As we always do, we track hot themes in the last 3 days on the home page and I noticed that one of the theme’s over the last 3 days was “yoo memo”. This theme referred to an 81 page memo authored by John C. Yoo that released to the public and discussed, in part, authorizing torture of government detainees. I certainly don’t want to give the impression that we are in the political analysis business by posting an opinion about politics, however I did find it interesting that there was a post about the lack of coverage of the memo from traditional media outlets. The fact that it was one of our hot themes in the political blogs, and hardly mentioned on the big news networks, says a lot about how lines are being drawn defining “newsworthy” information.
Apply that to business and imagine how difficult it may be for companies or businesses to know what is being discussed in the blogosphere if they only focus on analyzing the traditional outlets. The hottest theme in the blogopshere may not be what is coming in on your news feeds - and while this was about a nationally released government memo - imagine what you could be missing out on regarding your company’s brands or products?
Disclosure of Torture Memo Fails to Grab Tradional Media’s Attention
- Christine Sierra's blog
- Login or register to post comments
Comments
Interesting analysis. Of course, as you point out, it’s not just a majority vote. 35% negative may have been high enough to justify pulling the ad.
In any case, I’m more interested in this whole episode as a wake-up call to the power of social media to affect real-world events. For more detail, check out:
http://thenoisychannel.com/2008/11/20/tweet-first-ask-questions-later/
Nice thought but this analysis is by nature biased. If you didn’t capture the tweets from the initial outcry then you’ve vastly under-represented those who were negative about the ad.
That’s what happens with a groundswell…initial reactions are to the ad itself…later reactions are to the initial reaction.
Good thought but it would have had to be executed from the start to be real data.
Jeninma,
We absolutely agree - we would have liked to have had access to all the tweets for real analysis, but are limited by Twitter. Full analysis - across tweets AND blogs would have offered a more robust look into the situation.
Christine
Hi,
Did you do a distribution of women to men, and mothers and fathers to people without children?
Since most tweets I saw from women were negative over the weekend, and mostly parents were negative, but lots and lots of tweets from very active types (mostly men without children) were positive or neutral, I’m wondering how the target audience felt.
I don’t care so much about the distribution overall. I do care about whether moms didn’t like the ad, and maybe dads (dads seem much less likely to take motrin themselves because of past marketing to PMS problems).
My views were based upon anecdotal evidence, but I don’t think your analysis above proves your case, without getting the numbers in your target audience for the ad.
On a side note, I thought the tweets were overblown when I first saw the moms getting upset (and a few dads) that I follow. Then I watched the ad. I was put off enough that I probably won’t buy motrin anymore and I don’t have kids. I just thought the low grade hostility toward parents was obnoxious, and the impression stays with me. I don’t particularly want to take something that in a veiled way seems to disregard the very important bond between parents and their children especially with an infant.
Mary,
Our information is strictly based on text from within the tweets - it is what our software does when measuring sentiment. So we could not come to any conclusion on gender. We are not indicating that we think Motrin was right or wrong in their decision, we are asking marketers to think about how they analyze the data and what their own threshholds are before making any decisions. It’s still a bit fuzzy on how to measure social media and we believe sentiment is a key piece of that puzzle.
Christine
Christine,
Understand. But that’s the problem with stats. They don’t always show what is important. If a million folks with no kids or interest in kids tweeted just so they could jump into yet another controversy, then you have a lot of data that doesn’t really matter to motrin.
I realize there isn’t a good way, without spending a lot of time on qualitative analysis to get what I’m looking for. But I think that is the data you want to look at if you are that company.
I’m also curious if social media users in the first place represent an audience that is unrepresentative of Motrin’s target audience, e.g., because they are more likely to use generic equivalents.
Just as a pair of data points, my wife and I, who both wear our baby (she more often than I), liked the ad, though it wasn’t enough to induce us to go out and buy Motrin. De gustibus non est disputandum.
I think Motrin caved to a few vocal people - what a shame. I wrote Motrin’s Ad Agency directly, but never posted my opinion publicly on Twitter because I couldn’t afford the backlash these ridiculous people stirred up. Silence on my part, and the part of others, does not equal agreement, it just means that we put ourselves, our business, and common sense first.
Remember that old saying, “there’s no such thing as bad publicity.”
What do you suppose the ad reaction has done for Motrim’s brand awareness?
What proportion of the negatives reached beyond the ad campaign and reflect lasting sentiment about Motrim the product? That is, there might be folks who say to themselves, “that was a stupid ad, but I do need a pain reliever so maybe I’ll try Motrim.”
These are significant questions.
Seth