Does Sentiment Analysis Hold The Secrets to “Surviving” a Scandal?

Does Sentiment Analysis Hold The Secrets to “Surviving” a Scandal?
By Asher Feldman, General Sentiment

In the always-on news environment today’s media thrives in, anything that can be labeled a “scandal,” and take up significant portions of any day’s cycle, is worth its weight in gold. So it’s no surprise when simple things, like a misfired tweet or a misplaced iPhone prototype is labeled as such.

When the scandal tag is truly appropriate, however, the power today’s social media holds is never more apparent. The tail of anything the news loves labeling with a “-gate” elongates when social outlets get a hold.

But like anything, people eventually lose interest and move on. The only question is when the skies will finally clear. It could be a week, it could be a month; you never can tell. It’s not enough though to wait for people to stop talking about it. You need to know when it’s safe to come out of the woodwork again. That’s where sentiment analysis can step in.

For advertisers looking for an opportunity to pounce on an endorser when they’re turning the corner on a scandal, for a politician who’s looking for the opportune time to reenter the political world, for a company looking to move past a faux pas of sorts — social media sentiment analysis can hold the key. Too soon and you could risk lingering resentment, too late and you could be too far out of the public consciousness.

Take, for instance, one of the juiciest scandals to hit politics in the last decade — former New York Rep. Anthony Weiner’s resignation and subsequent attempts to return to politics. From May 27, 2011, the day he sent his ill-advised tweet, to September 2011, when the election to replace him was held, Weiner’s average sentiment rarely crept into positive territory, averaging -30 on a scale from -100 to +100. Understandably, the sky was dark around Weiner.

From that September until April 2013, when Weiner made clear his intentions to return to politics to The New York Times, Weiner’s sentiment continued to suffer, with a -26 daily average over the 2.5-year span. But things were improving. Over the next month, leading up until the day he announced his candidacy for the Democratic nomination for the 2013 NYC mayoral race, Weiner’s average sentiment score improved to -19, not positive, but better than it had been. But his scandal still loomed large.

Though his May 21 announcement juiced sentiment and volume counts for Weiner, things never looked truly positive for a Weiner candidacy. Had the “Carlos Danger” portion of the mayoral race never happened, Weiner might have been able to overcome his past, but with a maximum monthly sentiment of just +2 in 2013, Weiner never stood much of a chance against the likes of Bill de Blasio (a +36 monthly sentiment maximum), or Christine Quinn (+25 monthly sentiment maximum) on social media.

Would there have been a better time to try to return to politics? It’s not entirely clear, but Weiner’s case study could be one for other politicians to turn to. Media was not over a lingering impression of the former Congressman, and went looking for more dirt, finding exactly what they needed to bury the already disliked figure.

Sports may offer one of the truly comparable stages to measure “scandal” to that of politicians. In recent years, Lance Armstrong, Tiger Woods, Alex Rodriguez and others have had to weather strong negativity, to different degrees of success. Woods has been able to effectively return to the sporting world while maintaining a smaller profile after his infidelity was revealed. Woods has also returned as the face of marketing campaigns. How has he accomplished this turnaround?

Woods, unlike Weiner, chose his fights wisely and can look to sentiment analysis to prove he’s taken the right approach to recovering from seemingly crushing media scrutiny. From an all-time low of -30 monthly sentiment for December 2009, Woods waited the scandal out, returning to golf only when his monthly sentiment had return to +4, in April 2010, and only returning to a truly normal schedule in February 2012, when his monthly sentiment score was a healthy +46.

Sponsors were back on board, television and media had forgiven, and Woods was all the way back.

Could sentiment analysis have played a part in these decisions if it was available to these scandal-ridden celebrities? It only could have helped. Perhaps A-Rod and Chris Christie can take a hint.
[starbox id=afeldman]


  1. Can you say more about the data analyzed here? Did you just search for “Anthony Weiner” and code those by hand? Just search for “Weiner” and use an automatic tool (how good is it at disambiguating…the Weiner of interest?). Is it all Twitter data? Thanks.